Advances in Parasitology, Vol. 66 [1st ed.] 978-0-12-374229-2

Citation preview

Advances in

PARASITOLOGY VOLUME

66

EDITORIAL BOARD M. COLUZZI Department of Public Health Sciences, Section of Parasitology ‘Ettore Biocca’ ‘Sapienza – Universita` di Roma’ P. le Aldo Moro, 5, 00185 Roma, Italia

C. COMBES Laboratoire de Biologie Animale, Universite´ de Perpignan, Centre de Biologie et d’Ecologie Tropicale et Me´diterrane´enne, Avenue de Villeneuve, 66860 Perpignan Cedex, France

D. D. DESPOMMIER Division of Tropical Medicine and Environmental Sciences, Department of Microbiology, Columbia University, 630 West 168th Street, New York, NY 10032, USA

J. J. SHAW Instituto de Cieˆncias Biome´dicas, Universidade de Sa˜o Paulo, av. Prof Lineu Prestes 1374, 05508-990, Cidade Universita´ria, Sa˜o Paulo, SP, Brazil

K. TANABE Laboratory of Malariology, International Research Center of Infectious Diseases. Research Institute for Microbial Diseases, Osaka University, 3-1 Yamada-Oka, Suita, 565-0871. Japan

Advances in

PARASITOLOGY VOLUME

66 Edited by

D. ROLLINSON Department of Zoology The Natural History Museum Cromwell Road, London, UK

S. I. HAY Senior Research Fellow Malaria Public Health & Epidemiology Group Centre for Geographic Medicine KEMRI/University of Oxford/Wellcome Trust Collaborative Programme, Nairobi, Kenya

AMSTERDAM • BOSTON • HEIDELBERG • LONDON NEW YORK • OXFORD • PARIS • SAN DIEGO SAN FRANCISCO • SINGAPORE • SYDNEY • TOKYO Academic Press is an imprint of Elsevier

Academic Press is an imprint of Elsevier 84 Theobald’s Road, London WC1X 8RR, UK Radarweg 29, PO Box 211, 1000 AE Amsterdam, The Netherlands The Boulevard, Langford Lane, Kidlington, Oxford OX5 1GB, UK 30 Corporate Drive, Suite 400, Burlington, MA 01803, USA 525 B Street, Suite 1900, San Diego, CA 92101-4495, USA First edition 2008 Copyright # 2008 Elsevier Ltd. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means electronic, mechanical, photocopying, recording or otherwise without the prior written permission of the publisher. Permissions may be sought directly from Elsevier’s Science & Technology Rights Department in Oxford, UK: phone (+44) (0) 1865 843830; fax (+44) (0) 1865853333; email: [email protected]. Alternatively you can submit your request online by visiting the Elsevier web site at http://elsevier.com/locate/permissions, and selecting Obtaining permission to use Elsevier material. Notice No responsibility is assumed by the publisher for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions or ideas contained in the material herein. Because of rapid advances in the medical sciences, in particular, independent verification of diagnoses and drug dosages should be made. ISBN: 978-0-12-374229-2 ISSN: 0065-308X For information on all Academic Press publications visit our website at books.elsevier.com Printed and bound in The Netherlands 07 08 09 10 11 10 9 8 7 6 5 4 3 2 1

CONTENTS

Contributors Preface Obituary

1. Strain Theory of Malaria: The First 50 Years

vii ix xi

1

F. Ellis McKenzie, David L. Smith, Wendy P. O’Meara, and Eleanor M. Riley Introduction Background Clinical Virulence Reaction to Anti-malarial Remedies Infectivity Antigenic Properties Latency and Relapse Summary and Discussion Acknowledgements References 1. 2. 3. 4. 5. 6. 7. 8.

2. Advances and Trends in the Molecular Systematics of Anisakid Nematodes, with Implications for their Evolutionary Ecology and Host–Parasite Co-evolutionary Processes

2 3 6 9 12 16 28 33 38 38

47

Simonetta Mattiucci and Giuseppe Nascetti Introduction Molecular Systematics of Anisakid Nematodes The Current Taxonomy Phylogenetic Systematics of Anisakid Nematodes Genetic Differentiation in Anisakids Host–Parasite Cophylogeny Host Preference, Ecological Niche and Competition Anisakids as Biological Indicators Conclusions and Identification of Gaps in Our Knowledge of Anisakids to be Filled by Future Research Acknowledgements References 1. 2. 3. 4. 5. 6. 7. 8. 9.

49 51 55 91 100 108 114 119 130 137 137

v

vi

Contents

3. Atopic Disorders and Parasitic Infections

149

Aditya Reddy and Bernard Fried Introduction Atopic Disorders Relationship of Parasites to Atopic Disorders Laboratory Studies on Atopy Using Selected Parasites and Rodent Models 5. Concluding Remarks References 1. 2. 3. 4.

4. Heartworm Disease in Animals and Humans

150 151 154 175 182 184

193

John W. McCall, Claudio Genchi, Laura H. Kramer, Jorge Guerrero, and Luigi Venco Introduction (Biology and Life Cycle) Epidemiology in Domestic and Wild Hosts Pathogenesis, Immunology and Wolbachia Endosymbiosis Canine Heartworm Disease Feline Heartworm Disease Heartworm Disease in Ferrets Human Dirofilariosis Emerging Strategies in Heartworm Treatment and Control Acknowledgments References 1. 2. 3. 4. 5. 6. 7. 8.

195 199 209 222 233 253 258 263 265 265

Index

287

Contents of Volumes in This Series

295

See Colour Plate Section in the back of this book

CONTRIBUTORS

Bernard Fried Department of Biology, Lafayette College, Easton, Pennsylvania 18042, USA. Claudio Genchi DIPAV, Sezione di patologia Generale e Parassitologia, Universita` degli Studi di Milano, Via Celoria 10, 20133 Milano, Italy. Jorge Guerrero Department of Pathobiology, School of Veterinary Medicine, University of Pennsylvania, Philadelphia, Pennsylvania 19104, USA. Laura H. Kramer Dipartimento di Produzione Animale, Universita` di Parma, via del Taglio 8, 43100 Parma, Italy. Simonetta Mattiucci Department of Public Health Sciences, Section of Parasitology, ‘‘Sapienza’’—University of Rome, P.le Aldo Moro, 5, 00185 Rome, Italy. John W. McCall Department of Infectious Diseases, College of Veterinary Medicine, University of Georgia, Athens, Georgia, 30602, USA. F. Ellis McKenzie Fogarty International Center, Building 16, National Institutes of Health, Bethesda, Maryland 20892, USA. Giuseppe Nascetti Department of Ecology and Sustainable Economic Development—Tuscia University—Via S. Giovanni Decollato, 1, 01100 Viterbo, Italy. Wendy P. O’Meara Fogarty International Center, Building 16, National Institutes of Health, Bethesda, Maryland 20892, USA. Aditya Reddy Department of Biology, Lafayette College, Easton, Pennsylvania 18042, USA.

vii

viii

Contributors

Eleanor M. Riley Department of Infectious and Tropical Diseases, London School of Hygiene and Tropical Medicine, London WC1 E 7HT, United Kingdom. David L. Smith Zoology Department and Emerging Pathogens Institute, 223 Bartram Hall, University of Florida, Gainesville, Florida 32611, USA. Luigi Venco Clinica Veterinaria Citta` di Pavia, Viale Cremona 179, 27100 Pavia, Italy.

PREFACE

Ellis McKenzie, of the Fogarty International Center at the National Institutes of Health, and colleagues open this volume with a review of the history of strain theory in malaria research. This comprehensive review looks back to the origins of the concept of ‘varieties, strains or races’ of the Plasmodium species that cause human disease. The historical perspective allows insight not only to how these ideas developed but also how they affected clinical practice and epidemiological study. The development of these theories is then explored in relation to several themes: parasite phenotypes related to clinical virulence, reactions to anti-malarial drugs, infectivity to mosquitoes, antigenic properties and host immunity, latency and relapse. The authors conclude by discussing how these definitions of strain have evolved in relation to discoveries around each of these themes and by commenting on where ambiguity in working definition of a malaria strain remains. The next paper concerns the molecular systematics of the anisakid nematodes of the genera Anisakis, Pseudoterranova and Contracaecum. Simonetta Mattiucci from the University of Rome and Giuseppe Nascetti from Tuscia University, Italy, draw on their wealth of experience on working with this group and present a detailed account of the current understanding relating to the relationships of the different species involved and the implications for evolutionary ecology. It has long been known that morphological characters alone are insufficient to differentiate the species and each genus includes a number of sibling species. Once good molecular markers are available, it is possible to examine questions relating to host specificity and distribution. Comparing host and parasite phylogenies can be a powerful way of determining co-divergence and host-switching events, and the authors provide interesting insights by comparing Anisakis and the cetaceans and Contracaecum and the pinnipeds. Aditya Reddy and Bernard Fried of the Department of Biology, Lafayette, Easton, USA provided an interesting overview of the relationship between atopic disorders and parasitic infections. This is a topic that attracts much interest, especially in relation to helminth infections, but the associated literature tends to be scattered in diverse journals. Both helminth infections and allergy are common diseases, and the general observation is that helminth infections tend to be negatively associated with atopy, prevalence of allergic diseases and the severity of asthma.

ix

x

Preface

This review explores the protective mechanisms against atopic disorders which may be associated with parasites. This volume concludes with a chapter reviewing the biology and life cycle of the parasite Dirofilaria immitis, the causative agent of dog and cat heartworm. The chapter is contributed by John W. McCall, of the College of Veterinary Medicine of the University of Georgia, and an international assembly of co-authors. Advances in the understanding of its global distribution and prevalence of heartworm are outlined and the pathogenesis and immunology of infection discussed. The current understanding of the potential role of the Wolbachia endosymbiont in inflammatory and immune responses are discussed, along with the antibiotic treatment of infected animals. A large part of the chapter reviews the clinical presentation, diagnosis, prevention, therapy and management of the disease in dog, cats and ferrets. There is also a discussion of heartworm infection in humans, with notes on other epizootic filarial infections, particularly D. repens in Europe. The chapter concludes by examining novel treatments and highlighting the potential role of tetracycline antibiotics in adulticidal therapy. D. ROLLINSON S. I. HAY

OBITUARY Ralph Muller (1933–2007)

Ralph Muller, who died on 11 October 2007 as a result of prostate cancer, was a co-editor of Advances in Parasitology for 29 years. In 1978, Ralph (and I) joined the late Professor Russell Lumsden, then at the London School of Hygiene and Tropical Medicine, as assistant editors when Professor Lumsden took over as senior editor following the death of the founder, Professor Ben Dawes. When Professor Lumsden himself retired in 1981, Ralph and I continued to edit the publication; in 1994, we were joined by Dr. David Rollinson, of the British Museum (Natural History) in London. xi

xii

Obituary

I retired from this role at the end of 2006, to be replaced by Dr. Simon Hay (University of Oxford, UK), but Ralph continued in harness until shortly before his death. Some years before becoming an editor, in 1971, Ralph had begun his association with Advances by contributing a lengthy article on dracunculiasis to volume 9 (pp. 73–151). Ralph was one of the foremost parasitologists of the twentieth century in the UK. Beginning his life’s work as a parasitologist in 1955, when he graduated in zoology from Queen Mary College (University of London), Ralph moved to King’s College (London) where he worked on the maintenance in vitro of Haplometra cylindracea, a trematode living in the lungs of frogs, thus obtaining his Doctor of Philosophy (PhD) degree in 1958. Ralph was awarded Fellowship of the Institute of Biology in 1972 and the degree of Doctor of Science by London University in 1989. Ralph continued to work in King’s College as a research fellow until 1960. He first obtained ‘hands on’ experience of parasitology in the tropics during the two years, from 1960 to 1962, as a scientific officer working for the UK Overseas Development Administration on the control of schistosomiasis. Although based in London, this work involved visits to Kenya, Tanzania and Uganda, all then part of British East Africa. Following this, Ralph crossed the continent to become lecturer and chief of the subdepartment of parasitology in the University of Ibadan in Nigeria from 1962 to 1966. Here he became acquainted with the guinea worm Dracunculus medinensis, in which he retained a life-long interest. In 1966, he joined the Department of Parasitology of the London School of Hygiene and Tropical Medicine, where his continued study of guinea worm infection helped to draw attention to the importance of this infection which, though not of itself fatal, was a cause of considerable morbidity in affected populations and could often be the portal of entry of various pathogenic bacteria, including Clostridium tetani. Ralph later put his experience of dracunculiasis to good use through his association with the World Health Organization’s global programme to eradicate guinea worm infection—a target achieved in Asia and many, but not yet all, African countries; the current target date (not the first!) for global eradication is 2009. After 15 years at the London School, Ralph left in 1981 to become director of the International Institute of Parasitology (IIP) in St Albans, Hertfordshire, in the UK (where Ralph also lived). The Institute was part of an inter-governmental organization founded as the Commonwealth Agricultural Bureaux (CAB), and later was renamed as CAB International. The institute’s major function was to disseminate information concerning helminth parasites, primarily through the abstracting journal Helminthological Abstracts. This journal was later joined by the complementary Protozoological Abstracts. IIP was also instrumental in setting up relevant research projects in tropical countries. Ralph remained in this

Obituary

xiii

post until 1993; to the regret of many parasitologists, the Institute was closed in 1998, though publication of the abstracting journals continues. Ralph served as a Council member of the Royal Society of Tropical Medicine and Hygiene from 1982 to 1985. From 1993, after his retirement from the IIP, until the end of his life, Ralph continued to put his wide knowledge of helminthology at the disposal of students as a Visiting Lecturer at Imperial College in London and an Honorary Senior Lecturer at the London School of Hygiene and Tropical Medicine. In addition to editing Advances in Parasitology for many years, Ralph edited the primary research publication Journal of Helminthology (originally published by the London School of Hygiene and Tropical Medicine, now by Cambridge University Press), a post which he filled from 1972 to 1980 and again from 1987 to 1995. Ralph was the author or co-author of well over 100 scientific papers and also of two books: Worms and Disease: A Manual of Medical Helminthology (William Heinemann, 1975) and Medical Parasitology, with myself as coauthor (J. P. Lippincott Company/Gower Medical Publishing, 1990). The former book, re-titled Worms and Human Disease and with additional material by Derek Wakelin, achieved a second edition (CABI Publishing, 2002). The later version regrettably was devoid of the first edition’s delightful frontispiece, a reproduction of a detailed drawing of a sorrowful case of multiple guinea worm infection, reprinted from a publication of 1674, Exercitationes de Vena Medinensis et de Vermiculis capillaribus infantium by G. H. Velschius (Augsburg). The inclusion of this frontispiece seems to me to embody two of Ralph’s many good qualities—his dry humour and also his wide-ranging interest in and beyond the basics of helminthology. While being the director of the IIP, Ralph also edited a computerized Bibliography of Onchocerciasis (1841–1985), published by CAB International in 1987. Having known Ralph, mainly as an editorial colleague on Advances in Parasitology, for some 30 years, like many others I shall greatly miss, as well as happily remember, him as a distinguished helminthologist and a true friend. Apart from his knowledge of helminthology, which was both broad and deep, I remember in particular his calmness, his good humour and his generally ‘laid back’ attitude, all of which were of great value in the occasional moments of hustle and crisis. These qualities he courageously continued to manifest throughout his final illness. In the 1970s and 1980s, Ralph and I made several visits to Libya as external examiners in the Medical School at Garyounis University in Benghazi and the Medical Faculty of Al Fatah University in Tripoli. Ralph’s sense of humour contributed greatly to my enjoyment of these visits. I remember his delight on one such visit when he found that, by bounding up the stairs to our shared hotel room, he would be just in time to watch the televised broadcast of a World Cup football international game in which England was participating; regrettably, being perhaps rather less

xiv

Obituary

FIGURE 1

Ralph Muller PhD, DSc, CBiol, FIBiol.

FIGURE 2 Ralph Muller in the field, somewhere in West Africa (photo copyright # CAB International; reproduced by permission).

Obituary

xv

interested in sporting events than Ralph (who was himself a keen soccer player), I cannot remember whether they won. He was somewhat less delighted to find, on another occasion, that we were expected to share not only a room but also a double bed. I should add that a second bed was hastily procured. I am very grateful to Barnaby and Harriet Muller for their helpfulness in providing me with some of the above information. JOHN BAKER

This page intentionally left blank

CHAPTER

1 Strain Theory of Malaria: The First 50 Years F. Ellis McKenzie,* David L. Smith,† Wendy P. O’Meara,* and Eleanor M. Riley‡

2 3 6 9 12 16 16 19 23 28 33 38 38

Contents

Introduction Background Clinical Virulence Reaction to Anti-malarial Remedies Infectivity Antigenic Properties 6.1. Homologous and heterologous response 6.2. Clinical and parasitological response 6.3. Superinfection 7. Latency and Relapse 8. Summary and Discussion Acknowledgements References

Abstract

From the 1920s to the 1970s, a large body of principles and evidence accumulated about the existence and character of ‘strains’ among the Plasmodium species responsible for human malaria. An extensive research literature examined the degree to which strains were autonomous, stable biological entities, distinguishable by clinical, epidemiological or other features, and how this knowledge could be used to benefit medical and public health practice. Strain theory in this era was based largely on parasite phenotypes related to

1. 2. 3. 4. 5. 6.

* Fogarty International Center, Building 16, National Institutes of Health, Bethesda, Maryland 20892, USA {

{

Zoology Department and Emerging Pathogens Institute, 223 Bartram Hall, University of Florida, Gainesville, Florida 32611, USA Department of Infectious and Tropical Diseases, London School of Hygiene and Tropical Medicine, London WC1 E 7HT, United Kingdom

Advances in Parasitology, Volume 66 ISSN 0065-308X, DOI: 10.1016/S0065-308X(08)00201-7

2008 Elsevier Ltd. All rights reserved.

#

1

2

F. Ellis McKenzie et al.

clinical virulence, reactions to anti-malarial drugs, infectivity to mosquitoes, antigenic properties and host immunity, latency and relapse. Here we review the search for a definition of ‘strain’, suggest how the data and discussion shaped current understandings of many aspects of malaria and sketch a number of specific connections with perspectives from the past 30 years.

1. INTRODUCTION In the early 1920s, as debates about the number and nature of species that cause human malaria were receding, the idea emerged that each of these species consists of ‘varieties, strains or races’. Over the next 50 years, a complex set of ideas developed about such entities—whether they were discrete, independent, mutable; what traits distinguished them; how they affected clinical and epidemiological observations and interventions. By the late 1970s, cloning and cultivation of P. falciparum were possible, serologic and molecular techniques were developing rapidly, and concepts emerging from the rise of laboratory-based studies were accompanied by a shift in language: ‘strain’ and ‘race’ were largely displaced by the new terms ‘clone’ and ‘isolate’. In the mid-1990s, the concept of a ‘strain’ reemerged in anticipation of a vaccine protective against a subset of circulating parasites. This ‘strain theory’ assumed that malaria comprised discrete, independently transmitted, immutable entities, and concluded that ‘control of malaria through vaccination may be far easier than previously assumed’ (Gupta and Day, 1994a). Related theory was invoked to explain immunity to clinical malaria (Gupta and Day, 1994b), and, more recently, to describe immunity to var gene products (Gatton and Cheng, 2004). Early investigations of ‘strains’ in malaria were devoted to understanding parasite phenotypes, but, as rapid and reliable molecular techniques for determining parasite genotypes were developed, it seemed clear that ‘strains’ could be distinguished with unprecedented precision, either directly or by mapping genotypes to some smaller set of phenotypes. To do so, however, would require at least a provisional definition of ‘strain’. Here we review the initial search for a definition of ‘strain’, as the theory developed in interplay with practice in the 1920s–1970s. That search is not only of historical interest, but is also relevant as contemporary studies begin to revisit and rediscover important aspects of parasite phenotypes. This historical review points to important themes for contemporary malariology, and reminds us that sorting and classifying the objects of study critically determines how research is framed and pursued. Theories do not simply catalogue observations: they formulate general principles explaining many specific observations in terms of relatively few underlying entities, forces and relationships, so the

Strain Theory of Malaria: The First 50 Years

3

evidence accumulating for or against a theory often informs a wider field. Thus the pursuit of ‘strains’ was integral to the understanding developed in the 1920s–1970s of many aspects of malaria, notably its immunology— the now-familiar distinction between clinical and parasitological aspects of response, for instance. Here, we quote from the original literature as much as possible, to emphasize the richness of the resource and the degree to which some key points have been echoed and others neglected in recent literature. In a box at the end of each main section, we sketch several such connections between historical perspectives and current knowledge.

2. BACKGROUND ‘Shall we not be obliged to say that tertian and quartan agues are divine too, for nothing can be more regular than the process of their recurrence? But all such phenomena call for rational explanation’ (Cicero, 45BC). For millennia, the ‘intermittent’ fevers associated with chills (‘agues’) were distinguished and classified on the basis of periodicity—quotidian (daily), tertian (every other day) or quartan (every third day). In 1880, Laveran described crescent-shaped parasites in the blood of a soldier suffering the fevers of ‘paludism’ (malaria). He later distinguished an additional 46 parasite forms, and over the next 40 years maintained that ‘the different forms in which the haematozoa of paludism present themselves belong to one and the same polymorphic parasite’ (Laveran, 1893). In the late 1880s, Golgi linked periodicity in malaria fevers to parasite replication cycles, showed that morphologically distinct parasites were responsible for tertian and quartan fevers, attributed quotidian fevers to the presence of ‘different parasite generations reaching complete development at a day’s interval’ and, while noting that ‘the numerous varieties of intermittent malarial fevers reported are, in the very great majority simple varieties or combinations of the two fundamental types (tertian and quartan fevers)’, posited a third distinct type, a tertian parasite characterized by crescent forms, quinine resistance and ‘irregular’ fevers (Golgi, 1889). In the 1890s, Marchiafava, Bignami, Celli and their colleagues argued that ‘there is no group of fevers which are naturally and per se irregular, but fevers of every class may become irregular, and in different ways’ (Marchiafava and Bignami, 1894). Like Golgi, and Laveran (with respect to ‘parasitic elements’), they recognized that many problems in interpretation arose from concurrent infections by different parasite species and ‘colonies’ (Golgi’s ‘generations’) within a species, given that fevers in some way followed the relative densities, transitions or transformations of parasites: ‘difficulty is met with in the study of the mixed malarial

4

F. Ellis McKenzie et al.

infections . . . it is easy to predict the possibility of a very large series of combinations, caused by the number of the parasitic colonies, by the way in which their life cycles are, so to say, interwoven, &c.; as a resultant therefrom, the fever takes the most different courses’ (Marchiafava and Bignami, 1894). But they insisted on the existence of four distinct entities, including a ‘true’ quotidian parasite. They differentiated springtime tertians and quartans from ‘malignant’ aestivo-autumnal (summer-fall) crescent producers, and in the latter distinguished tertians and quotidians: ‘The spring tertian and quartan parasites give rise to the mild forms; the aestivo-autumnal tertian and, more rarely, aestivo-autumnal quotidian . . . give rise to the severe forms’ (Celli, 1901). In 1910, Ross who had demonstrated that malaria parasites are transmitted by mosquitoes, wrote of ‘three different types of fever, the quartan, the tertian, and the irregular or malignant type . . . Since the time of Golgi, all observers admit . . . that these three types are different species . . . Some authors consider that there are two if not three varieties (or ? species) of malignant parasites. I am inclined to agree with them, but have not yet satisfied myself sufficiently on the point to admit it in my classification’ (Ross, 1910). He too noted that some patients had concurrent infections with different parasite species, each of which might consist of ‘two or three sets sporulating [i.e. completing schizogony] on different days . . . The rule generally accepted is that each set of parasites continues it own evolution independently of other sets which may be present. But much more precise work requires to be done on this point’. The ‘mild’ tertian and quartan parasites are now known as Plasmodium vivax and Plasmodium malariae, respectively, the ‘malignant’ tertians as Plasmodium falciparum (‘falciform’ means ‘crescent-shaped’, a reference to the distinctive gametocytes—the forms transmissible to mosquitoes—of P. falciparum). Quotidian and irregular fevers are rarely mentioned in recent literature, but would be attributed to double tertian, triple quartan or other combinations of infections. The last of the four species known to cause malaria infections in humans in nature, the ‘tertian’ Plasmodium ovale, was fully described in 1922 (Stephens, 1922). In that same year, citing several observations in the literature—notably on the low gametocyte production in P. falciparum infections on the coast of West Africa, compared to those in Italy and Macedonia, and on the development of clinical malaria in ‘resistant’ West Africans, who moved from the coast inland, or between countries— Marchoux made a speculative leap: ‘There are not only 3 species or varieties of Plasmodium, but 3 groups within each of which there exist a considerable number of varieties’ (Marchoux, 1922). Laveran, Golgi, Marchiafava, Bignami, Celli, Ross and their contemporaries were familiar with experimental transfers of malaria parasites by blood inoculation, and had noted, for instance, that ‘one species of

Strain Theory of Malaria: The First 50 Years

5

parasite often disappears from the blood upon inoculation with a different species’ (Deaderick, 1909). Wagner-Jauregg observed that patients with severe syphilis were sometimes ‘healed through intercurrent infectious diseases’, which led him to ‘intentionally imitate this experiment of nature’ through blood inoculations of malaria (WagnerJauregg, 1922), which in turn led to his 1927 Nobel Prize. Before the advent of effective antibiotics, tens of thousands of neurosyphilis patients were treated in malariatherapy clinics in the United Kingdom, the United States, continental Europe and elsewhere. A 1984 review concluded that ‘malariatherapy was less expensive and produced clinical improvements more frequently and more rapidly than did the best drug treatments’ (Chernin, 1984). Contemporaneous descriptions of procedures are given in Badenski (1936), de Rudolf (1927), Kupper (1939) and Lomholt (1944). A recent, independent, explicit analysis of relevant ethical issues accompanies extensive information about patient participation and treatment, and descriptions of procedures used in two of the major US facilities, in Collins and Jeffery (1999). Parasites were transferred to patients by mosquito bite or by blood inoculation. Because most experts thought that the probability of the neurosyphilis cure was related in some way to the number and intensity of fevers, clinic staff often gave subcurative doses of anti-malarial drugs to manage infections accordingly, and then full therapeutic doses at the end of the treatment regimen. Patients not cured of syphilis by initial infections were sometimes re-infected, typically with some variation in material or procedure. As clinics tried new methods and combinations, their results appeared in research papers and at conferences; much of our current knowledge of malaria is founded on these observations. With respect to Marchoux’s ‘varieties’ and their classification, the practice of malariatherapy meant that ‘now strains of widely diverse origins can be brought together and inoculated into patients lying side by side in the same hospital’ (Hackett, 1937). In the 1920s and 1930s, the results reported from malariatherapy, in conjunction with several influential malaria experiments in birds and non-human primates, and field observations in malaria-endemic countries, produced ‘general agreement that within each species of malaria parasite there are races or strains which can be recognised as being distinct by their clinical virulence, their infectivity, their reaction to antimalarial remedies and their antigenic properties’ (James and Ciuca, 1938). This characterization of ‘races’ or ‘strains’ had a strong practical bent: ‘clinical virulence’ and ‘reaction to antimalarial remedies’ were important because therapeutic malaria infections should produce fevers and other symptoms at sufficient but not excessive levels, and should be manageable; ‘infectivity’ and ‘antigenic properties’ were important because parasite transfer and patient re-infection should be reliable and

6

F. Ellis McKenzie et al.

predictable. It soon became clear that ‘the latency and relapse patterns of malarial infections vary with the type and geographic origin of the specific parasites’ (Schwartz et al., 1950), and that this fifth distinguishing characteristic had a similar practical importance. In this paper, we use these five characteristics—clinical virulence, reaction to anti-malarial remedies, infectivity, antigenic properties, and latency and relapse—as a framework for reviewing the development of strain theory from the 1920s to 1970s.

3. CLINICAL VIRULENCE By the 1920s, it was ‘well known that infections vary greatly in severity, and that in some localities nearly every infection is pernicious, while in others pernicious symptoms do not develop, although the same plasmodium is presumably the cause of both’ (Craig, 1926). In India, for instance, ‘malarial fevers are usually most severe and persistent in low-lying coast districts’ (Hehir, 1927). If differences between parasites were responsible for ‘the common observation that in some parts of the world malignant malaria is more malignant than in others’ (James et al., 1932), then, for malariatherapy, ‘pure tested strains, if available for inoculation, are far and away preferable, since not all strains are equally suitable. Some cause only quite mild reactions, whilst others produce severe reactions with more pronounced clinical symptoms’ (Nocht and Mayer, 1937). It was also thought ‘possible that different strains of Plasmodium vivax give different clinical types of malaria’ (Grant, 1923), and that ‘the tendency also of P. falciparum to select certain organs—the intestinal wall, the omentum, the brain—for localization has been explained by tropisms developed by certain strains’ (Hackett, 1937). Though differences in virulence were often said to be dramatic, the differences were seldom specified in terms of symptoms, and, when specified—for example, ‘the strains under review differed in regard to manner of onset, character of the fever and highest pulse-rates’ (de Rudolf, 1924)—were seldom quantified. The clinical differences most often specified and quantified between strains related to fevers—‘a very marked difference between the two with respect to the height of temperature produced’, for example (Bunker and Kirby, 1925). Later, in the 1940s, the periodicity of fevers with the Baltimore, St. Elizabeth and New Hebrides strains of P. vivax was reported as 41.5, 43.4 and 45.8 h, respectively, which ‘suggests that each strain might have a characteristic periodicity . . . [which] will be a valuable point in distinguishing strains’ (Young, 1944). Strains were typically identified by their geographic site or clinic of origin, the former sometimes assumed rather than known. For instance, Horton Hospital, in England, most often used the ‘Madagascar’ strain of P. vivax, obtained in 1925 from an Indian seaman at an English port

Strain Theory of Malaria: The First 50 Years

7

and named for the putative source of his infection. Compared to a ‘Holland’ strain of P. vivax judged inappropriate for malariatherapy, the Madagascar strain had a shorter incubation period (interval between inoculation and first fever), a lower frequency of spontaneous recovery and higher frequencies of quotidian fever and fevers above 106.4oF (James and Ciuca, 1938). With P. falciparum, Horton Hospital staff found that ‘the Italian and Sardinian strains are more severe than any we have obtained from India and Africa’, based largely on their observation that ‘severe clinical symptoms’ developed much more rapidly after inoculation. Roumanian strains of P. falciparum were found to be less severe than the Italian-Sardinian, but more severe than the Indian–African (Shute and Maryon, 1954). Another ‘means of comparing the severity of the cases’ at Horton Hospital was the ‘longest period of fever without a fall to the normal temperature’, which averaged 74 h with the P. falciparum Rome I and Sardinia strains, 37 h with the Indian I and Indian II strains. Further evidence of these ‘striking clinical differences’ was that the Italian strains required eight times as much quinine for treatment as did the Indian strains, on average, and ‘continued to relapse for a much longer time’ (James et al., 1932). The application of these studies was seen as urgent and direct: ‘physicians in all malarious parts of the world should endeavour as soon as possible to add to existing information on the clinical virulence of the particular strains of P. falciparum prevalent in the countries where they work. Are all the strains in India as mild as those reported in the paper and are all the Italian strains as virulent? If the patient is able to say in what place he contracted the infection, shall we have at hand information indicating what will be the probable course of his illness?’ (James et al., 1932). As the comparisons based on quinine and ‘relapses’ suggest, however, studies of differing virulence were entangled with questions about the nature of parasite drug response and latency, the role of the host and many other factors under investigation, including ‘whether the terms ‘‘immunity’’, ‘‘virulence’’, ‘‘infectivity’’, ‘‘epidemic strains’’, etc. current in work on immunology as it relates to bacterial infections should be used in work on the immunology of malaria’ (James and Ciuca, 1938). This succession of quotes from the Horton Hospital points to the interesting context of their use of the term ‘virulence’ during the 1930s. They maintained that ‘within the species there are various geographical races which . . . can be recognized as being distinct by their clinical virulence’ (James et al., 1932), and that ‘we have no evidence that the inherent virulence of a particular strain can be increased or diminished’ (James et al., 1936). However, they also revealed that ‘at Horton between 1925 and 1930 we succeeded in increasing the physical vigour and activity of an endemic strain of P. vivax from Madagascar to the degree in which it

8

F. Ellis McKenzie et al.

caused this severe ‘‘epidemic type’’ of the disease in 80% of our cases’ (James et al., 1936). The procedure was not described (see Shute, 1951 for a schematic), but ‘careful selection of donors and recipients’ (James and Ciuca, 1938) had produced ‘a strain which multiplies freely and vigorously’ and thereby increased severity because, all else equal, ‘the number of parasites govern the severity of the case’ (James et al., 1936). This increase in severity was ‘not an increase in virulence’, however, given ‘that ‘‘virulence’’ implied the possession or elaboration of a poisonous or venomous active principle (‘‘toxin’’). If so, it was not a good word to apply to the malaria parasite in which . . . no ‘‘toxin’’ had been found’ (James et al., 1936). Furthermore, it had become ‘evident that ‘‘infectivity’’, ‘‘virulence’’, and so on are relative terms depending as much on the receptivity of the host as on the biological properties of the parasite’ (James and Ciuca, 1938). Thus, at Horton Hospital, it appeared that ‘virulence’ might refer to an interaction between a particular host and a fixed property of a strain, in the form of a toxin, but that the clinical severity of a malaria infection might be determined by malleable as well as ‘inherent’ properties of the infecting strain. The contention that clinical features of a strain could change was not unique to Horton Hospital; however, the claim that ‘virulence tends to be increased with each successive transmission’ (Macbride, 1924) had been supported (Bunker and Kirby, 1925; Grant, 1923), contradicted (Fiertz, 1926; Yorke and Macfie, 1924) and confounded by mixed results (Lilly, 1925) in the mid-1920s. Furthermore, Wagner-Jauregg not only held the unnatural transmission [via blood inoculation] in malariatherapy responsible for ‘mildness’ and drug sensitivity, but seemed to associate the two traits: ‘inoculated malaria showed itself much more sensitive toward quinine than the natural [mosquito-transmitted] malaria . . . The mildness of this inoculation malaria may be explained thus, that the plasmodia which always reproduce themselves only in the asexual way are less capable of resistance’ (Wagner-Jauregg, 1922). Thus—despite early speculation ‘that a race of malarial parasites that is immune to quinine may be developed [as] fresh water amoebae may be gradually habituated to salt water’. (Leslie, 1910)—in the 1930s parasite response to a therapeutic drug was often taken as a token or reflection of ‘virulence’ as well as a marker of strain identity: ‘The virulence of any strain of parasite may be manifested by the toxic symptoms it produces in the host, and this influences the degree of ease or difficulty with which a clinical cure can be produced . . . but it may also be manifest by the power of the parasite to resist such means and to survive in the host’ (Sinton, 1931). At Horton Hospital, and elsewhere, the importance of distinguishing between clinical-severity and drug-response traits emerged gradually through the 1940s: ‘Failure of a strain of parasite to respond to a given drug is not, however, in itself evidence of virulence; it may be that the

Strain Theory of Malaria: The First 50 Years

9

strain, though resistant to this particular drug, is highly susceptible to others. The virulence or toxicity of a strain can best be seen by its effect on the host in the absence of drug treatment’ (Shute and Maryon, 1954). The search for a definition of virulence and the underlying causes continues, but ‘virulence’ has been used to describe so many different aspects of malaria that a specific operational definition remains elusive. The same historical difficulties and ambiguities dominate the current search, including the mutability of ‘virulence’ and the role of host factors in determining disease presentation. Because it is now considered unethical to allow experimental infections to progress to severe clinical symptoms, and because natural infections are difficult to observe, definitions of virulence have moved away from detailed clinical observations to focus on severe manifestations such as cerebral malaria and life-threatening anaemia. Although, selection studies in a laboratory rodent-malaria model imply that virulence (as indicated by host weight loss) is a heritable trait and may involve parasite replication rate (Mackinnon and Read, 1999), there are still no robust molecular markers of parasite ‘virulence’. Genetic association studies have failed to find any reproducible markers of parasites causing cerebral malaria or severe anaemia (A-Elbasit et al., 2007; Dobano et al., 2007), although some differences in the distribution of var gene alleles have been reported between severe malaria cases and controls (Bull et al., 2005; Kyriacou et al., 2006). Parasites sequestering in placenta and giving rise to pregnancy-associated malaria do share a common phenotype (adherence of infected red blood cells to chondroitin sulphate A via PfEMP-1) and some candidate var genes responsible for this phenotype have been identified (Francis et al., 2007; Salanti et al., 2004; Viebig et al., 2005).

4. REACTION TO ANTI-MALARIAL REMEDIES Uncertainty about drug resistance as a stable strain characteristic was evident in the extensive studies of anti-malarial drugs conducted during World War II: ‘there are two main types of strain—one relatively insusceptible to atebrin [mepacrine] and the other normally susceptible . . . [but] variation in the degree of susceptibility to atebrin amongst the relatively insusceptible strains . . . suggests that there may have been several strains of P. falciparum occurring naturally in the Aitaipe-Wewak area or that the phenomenon of atebrin insusceptibility was not so much an inherent characteristic of the strains as one which had been, or was, in the process of being acquired’ (Fairley, 1946). The initial description of the

10

F. Ellis McKenzie et al.

Chesson strain of P. vivax, also from New Guinea, also implies doubt about relying on drug response alone for strain definition: it ‘reacted differently to certain drugs than did the St. Elizabeth strain . . . This and other characteristics suggest that it might be a strain distinct from some of the American malarias’ (Ehrman et al., 1945). In 1949, using a strain of P. vivax from Hong Kong and a strain of P. falciparum from West Africa, ‘resistance to proguanil was produced . . . by giving small doses of the drug to consecutive patients in a series in whom the strain was being maintained by blood-inoculation’ (Seaton and Adams, 1949; Seaton and Lourie, 1949). Nonetheless, ‘the accidental production in the field of proguanil-resistant strains . . . is thought unlikely’, because high-level resistance had taken 20 months to arise in P. vivax and proguanil had a ‘sterilizing effect’ on P. falciparum gametocytes. At exactly the same time, proguanil resistance was detected in the field, in Malaya: ‘the widespread use of proguanil in the Tampin district for the past two years . . . we believe, is related to the emergence of resistant strains of P. falciparum’ (Edeson and Field, 1950). The authors noted that Rollo et al. (1948), based on avian malaria experiments, had offered the ‘likely explanation . . . that resistant parasites occur spontaneously as rare mutants in a normally sensitive strain; and, as a result of selective survival of these resistant mutants when the parasites are exposed to the drug, a stable resistant strain emerges’. One of the proguanil-resistant strains of P. falciparum from Malaya was later reported to be cross-resistant to pyrimethamine (Robertson et al., 1952), and another not (Davey and Robertson, 1957). The appearance and spread of pyrimethamine resistance in P. falciparum was demonstrated in field experiments in the mid-1950s in East Africa—‘during the course of the treatments the resistant varieties spread, replacing sensitive strains; upon cessation of the treatments resistant varieties regressed and became submerged’ (Clyde and Shute, 1954)—but whether their rise and fall occurred through competitive interactions or through ‘conjugation with other strains . . . could not readily be determined’. Macdonald’s summary of the situation, in 1957, was that ‘the reactions of strains of parasite are unpredictable and in consequence drugs enjoy very different reputations in various parts of the world’ (Macdonald, 1957). Soon after chloroquine resistance was detected in the early 1960s, in Colombia (Moore and Lanier, 1961) and Thailand (Young et al., 1963), a series of studies began to compare and differentiate ‘strains of Plasmodium falciparum resistant not only to chloroquine but to other groups of synthetic anti-malarial drugs’ on the basis of cross-resistance patterns, determining, for instance, that ‘the Malayan I strain differs from the Thailand strain in its susceptibility to pyrimethamine; the Malayan III strain differs in its susceptibility to mepacrine’ (Contacos et al., 1963).

Strain Theory of Malaria: The First 50 Years

11

Though all strains appeared susceptible to quinine, an experimental transfer of a Malayan chloroquine-resistant strain between quininetreated prison volunteers gave ‘strong evidence indicating the emergence of a variant strain characterized by lessened sensitivity to quinine compared to that of the parent strain’ (McNamara et al., 1967), an ‘exceptional’ result cited in support of the hypothesis that ‘new strains may continually be emerging in the endemic area and adapting to the intermittent and irregular drug pressure encountered there’ (Clyde et al., 1969). The World Health Organization’s conclusion, in 1969, was that ‘most of the drugs commonly used to treat malaria have shown variations in effect that are related to the strain of parasite . . . [but] changes in sensitivity may be transient or permanent. . . . The problem is particularly complex if one attempts to distinguish between the inherent differences in the response to drugs of various strains of the same species and the changing pattern of response induced by previous contact with drugs’ (WHO, 1969). In the 1930s, differences in drug response had already suggested that ‘each strain of tertian or of aestivo-autumnal malaria is a problem in therapeutics by itself’ (Hackett, 1937), and that ‘isolation, selection, and other factors known to bring about change in other organisms would act upon malarial parasites . . . [though] malariologists have been slow in acknowledging this possibility’ (Huff, 1938). By the 1980s, the challenge had become that ‘isolates that exhibit multiple drug resistance may, in fact, be mixed infections of parasites exhibiting resistance to each drug separately’ (Rosario, 1981), and the role of drug response in characterizing ‘strains’ had become less clear. The determination of the molecular/genetic basis of resistance to several anti-malarial drugs (Gregson and Plowe, 2005; Wellems and Plowe, 2001) has allowed many of these hypotheses to be tested. For example, a single point mutation (pfcrt K76T) provides high-level chloroquine resistance in P. falciparum infections, but the mutation is found with nine other mutations. Discrete point mutations in a single bi-functional molecule confer resistance to pyrimethamine and to sulphadoxine. These mutations occur with relatively high frequency but impose real fitness costs on the parasite that are observed when selection pressure is relaxed. Following the switch away from sulphadoxine–pyrimethamine as a first-line anti-malarial in East Africa, the regression of resistant varieties first described by Clyde and Shute (1954) and the re-emergence of sensitive strains has been documented (Hastings and Donnelly, 2005), confirming the partial loss of fitness associated with drug-resistance mutations. In Malawi, the loss of chloroquine resistance has been dramatic (Kublin et al., 2003). As postulated by Rollo et al. (1948), mutant parasites are selected and increase in frequency by spreading among

12

F. Ellis McKenzie et al.

hosts in populations exposed to drugs (Anderson and Roper, 2005; Pearce et al., 2005; Mita et al., 2006), but ‘conjugation with other strains’ (Clyde and Shute, 1954), that is, recombination, allows highly- or multiply-drug-resistant strains (carrying multiple point mutations) to emerge and allows spread into many different parasite genetic backgrounds. Thus, in the case of resistance to anti-malarial drug resistance, there appears to be a simple map between genotype and phenotype, and there is a clear basis for defining resistant strains.

5. INFECTIVITY Whatever their other characteristics, parasite strains in the field and in most malariatherapy clinics had to be transmissible from humans to mosquitoes (and back to humans) to persist. Gametocytemia was taken as a marker of transmissibility to mosquitoes, and gametocyte production seemed to vary between strains. At the Horton Hospital, for instance, the Roumanian P. vivax Apostol ‘strain was unsatisfactory because it did not produce a sufficient number of gametocytes’, in contrast to the ‘large number of gametocytes which are usually produced by this [Madagascar] strain’ (Shute, 1937). Dramatic changes in gametocyte production were associated with transfer solely by blood inoculation: ‘malaria strains when inoculated from patient to patient more or less rapidly lose the capacity of producing gametes, and thus of infecting mosquitoes’ (Pijper and Russell, 1926), and, though such ‘gametocyteless’ strains were sometimes propagated (Jeffery, 1951), clinics were cautioned to ‘keep patients undergoing malariatherapy in screened rooms’ because ‘one can never be sure that such loss of power to produce sexual forms is permanent, and strains of malaria are known which retain the ability to infect mosquitoes more or less indefinitely’ (Russell et al., 1946). While studies increasingly indicated that ‘gametocyte density is not a reliable guide to the probably resulting qualitative infection of mosquitoes’ (Boyd, 1942a), it was the most reliable gametocyte producers that became the strains commonly used in most clinics. As for the actual infection of mosquitoes, Anopheles species often showed differing susceptibilities not only to different Plasmodium species (Boyd and Stratman-Thomas, 1934; Boyd et al., 1935), but also to different strains within each species: ‘we failed entirely to infect our English maculopennis with the Indian strains of P. falciparum [but] with the European strains from Rome and Sardinia we succeeded . . . the capacity of each strain to infect a named species of anopheles must be separately worked out’ (James et al., 1932). Further studies with ‘English maculopennis’ at

Strain Theory of Malaria: The First 50 Years

13

Horton Hospital reported successes with P. falciparum strains from Roumania, and failures with strains from East and West Africa (Shute, 1940). European vectors transported to Africa gave a similar result: ‘experiments in Garki, Nigeria, with A. atroparvus from Italy have shown a refractoriness to infection with the local strain of P. falciparum’ (Ramsdale and Coluzzi, 1975). However, ‘A. gambiae, originating from the Nyanza province of Kenya, was able to transmit the Malayan strains of P. falciparum with no more difficulty than African strains’ (Davey and Robertson, 1957). Accordingly, even the broadest of summaries mentioned exceptions: ‘European strains of P. falciparum require a European mosquito to transmit them, and tropical strains tropical mosquitoes, but in the latter case it may be noted that the Indian A. stephensi is a good vector of the far distant West African P. falciparum’ (Garnham, 1966). In the United States, it appeared that A. quadrimaculatus and A. punctipennis ‘vary widely in their susceptibility to different strains of P. falciparum [but] are approximately equally susceptible to both strains of P. vivax’, and, as one result of such findings, the Rockefeller Foundation clinic ‘abandoned the propagation of the Coker strain’ but continued the Holzendorf, Long and Manuel strains of P. falciparum (Boyd and Kitchen, 1936a). As results of vector–parasite transmission tests accumulated, it appeared that ‘local strains of parasites may or may not show a high degree of adaptation to anophelines which are coindigenous to their own faunal regions, and . . . anophelines may or may not show a high degree of susceptibility to exotic strains of parasite’ (Boyd, 1940a), and the practical implications of strain infectivity were seen to extend well beyond malariatherapy clinics. At Horton Hospital, ‘results indicate that persons carrying gametocytes of P. falciparum of tropical origin would be unlikely to cause any outbreak of fresh cases of malaria in this country through the agency of our English maculopennis. On the other hand, our four English anophelines become infected when fed on tropical strains of P. vivax and our A. maculopennis is a very efficient carrier of P. ovale’ (Shute, 1940). In the United States, extensive transmission experiments with indigenous species of Anopheles and ‘exotic’ strains of P. vivax and P. falciparum were undertaken during World War II (e.g. Boyd, 1949) in part because ‘studying foreign malaria imported by returning servicemen, it has been shown that these strains are infective to and can be transmitted by our native malaria vectors’ (Young et al., 1947). A particular concern was that ‘returning soldiers with such infections may be responsible for the establishment in this country of epidemic or endemic foci for imported vivax strains’ (Watson, 1945). Extensive studies of P. vivax infections from the ‘Solomon Islands, New Hebrides islands, New Guinea, Tunisia, Liberia, Trinidad, and the China-Burma-India theater’ with the major malaria vectors of the Western

14

F. Ellis McKenzie et al.

and Eastern United States found that, based on the fraction of fed mosquitoes infected, ‘A.m. freeborni is more susceptible to these foreign malarias than A. quadrimaculatus’ (Young and Burgess, 1948), the same result as with ‘a domestic strain of P. vivax’, St. Elizabeth (Burgess and Young, 1950). P. vivax strains from the Mediterranean and India, however, infected a lower percentage of A. freeborni than A. quadrimaculatus (Young et al., 1949). Similar studies with P. falciparum showed that the Malayan IV strain was much more infective, and the domestic McLendon strain slightly more infective to A. freeborni than to A. quadrimaculatus (Coatney et al., 1971), that A. albimanus was highly susceptible to the Panama strain, but almost refractory to the Thailand strain (Collins et al., 1963), and so forth, thus that ‘the infectivity of isolates of P. falciparum to different anophelines is dependent to some extent on the geographical origin of either the parasite or the mosquito’ (Coatney et al., 1971). The usual inference was that, while ‘there may exist, between particular strains of parasites and their definitive hosts, a very high degree of local adaptation, which under certain conditions may conceivably be a natural barrier to the extension of the range of a given strain of the parasite . . . the human intermediate host is not likely to be a factor in limiting the extension of the range of strains of these parasites’ (Boyd et al., 1938a). In what was apparently the only published study of its kind, the results suggested that ‘anophelines infected on a person concurrently infected with two strains of a single species (P. vivax) may either: (a) become infected with but one of the strains, or (b) may possibly become infected with both of the strains and simultaneously propagate both’ (Boyd et al., 1941). Thus, it was plausible that ‘the happy adjustment of parasite to vector in any area has come about through a weeding-out process, in which ill-adapted strains (and species) have failed to be transferred at proper intervals and have consequently become extinct’ (Hackett, 1937). If there was any correlation between infectivity and gametocytemia, however, it seemed certain that ‘the accidental introduction of a strain of parasite producing large numbers of gametocytes would lead to an increase in incidence’ (Bishop, 1955), and, since ‘gametocyte output is considerably greater in European strains than in Indian or African strains’ of P. falciparum (Shute and Maryon, 1954), for instance, it was not clear what, other than strict local matching of vectors and parasites, might constrain such introductions, or selection for increased gametocyte production. This question led directly to another: ‘are the male gametocytes of one strain of a species of parasite able to fertilize the females of a different strain? . . . inability to do so would be essential if strains are to retain their identity where more than one occurs in a given locality’ (Shute and Maryon, 1954). If cross-fertilization were possible, ‘‘‘How long would an imported strain

Strain Theory of Malaria: The First 50 Years

15

retain its identity when it was introduced into an endemic area?’’ And the answer, we think, would be: ‘‘not very long providing the indigenous anopheles were susceptible to infection’’. . . . [This] helps to explain why it is fairly obvious that there are separate geographical strains, but so very much less obvious that there are separate ‘locality strains’ (Shute and Maryon, 1954). By the 1980s, in vitro cultivation of P. falciparum made it possible to distinguish between a ‘clone’ (‘genetically identical organisms derived from a single cell by asexual division’; Walliker, 1983) and an ‘isolate’ (‘a sample of parasites, not necessarily genetically homogeneous, collected from a naturally infected host on a single occasion’; Walliker, 1983). Studies soon demonstrated that gametocyte production and infectivity might differ between clones grown from a single isolate (Burkot et al., 1984; Graves et al., 1984a), and that ‘for those strains that appear to lose gametocyte formation in culture this is a result of selection operating on a mixed population’ (Bhasin and Trager, 1984). Indeed, inability of cultured asexual forms to produce gametocytes was found to be due to complete and irreversible loss of large segments of genetic material (Day et al., 1993). Host immunity that reduces gametocyte densities and that blocks infection in mosquitoes suggests a role for the host in determining infectivity. High gametocyte production in response to sulphadoxine–pyrimethamine suggests that gametocyte production can respond to some cues within the host, and is not a fixed trait (Barnes and White, 2005). Incompatibility between parasite strains and various species and strains of Anopheles is now believed to reflect an interplay between polymorphic innate immune response genes of the mosquito that promote or retard parasite killing (melanization) and unknown melanization resistance genes of the parasite (Vlachou and Kafatos, 2005), hence current discussions often invoke evolutionary costs (Lambrechts et al., 2005; Sinden et al., 2004). Simultaneous propagation of multiple strains of malaria by a single mosquito has been confirmed by molecular genetic studies (Babiker et al., 1994; Huber et al., 1998; Ranford-Cartwright et al., 1993), but the high frequency of recombination during the sexual phase of the parasite life-cycle (confirming that ‘male gametocytes of one strain’ can indeed ‘fertilize the females of a different strain’) may mean that outcrossing, between strains, is as common as ‘selfing’ and thus, as predicted by Shute and Maryon (1954), that strains rapidly lose their identity in endemic areas. Hence, as with drug response, the role of infectivity in maintaining and characterizing strains began to appear highly complex.

16

F. Ellis McKenzie et al.

6. ANTIGENIC PROPERTIES 6.1. Homologous and heterologous response In 1898, an army doctor observed that ‘the contrast between West Indians and the natives composing the Sierra Leone Frontier Police when serving in the field in the same column was most marked: the former much troubled by fever, the latter having none’, and inferred ‘that the malarial organism of the West Coast is of a different variety to the plasmodium of the West Indies, and that immunity acquired to the first mentioned confers no protection from the second’ (Smith, 1898). In India, ‘there is always a certain amount of malaria of local origin in most towns, but this affects mainly children and new-comers; it is probable that adults sometimes acquire a certain measure of immunity to local infection, but are not immune to the more virulent type found at certain places in the interior of the districts’ (Kenrick, 1910). By the 1920s, it was recognized that ‘immunity developed against one species of Plasmodium does not confer a similar protection against the other species’ (Yorke and Macfie, 1924) and, retrospectively, with respect to strains, that in the ‘intermingling of men coming from malarial regions’ during and soon after World War I, ‘inhabitants more or less resistant to local strains showed little or no immunity to foreign strains and the immigrants had no resistance against the local virus [parasite]’ (League of Nations Health Organisation Malaria Commission, 1934). With such intermingling, and ‘numerous bites from innumerable mosquitoes . . . the chance of a person receiving an infection with one or more strains of parasite of varying degrees of virulence was considerable . . .. [and] with infection by each new strain the sufferer would be liable to a recurrence of clinical manifestations of greater or lesser intensity’ (Sinton, 1931). Thus, it began to appear that ‘persons can easily be immunized against a particular strain . . . but that the resistance breaks down if they are inoculated . . . even with a different strain of the same species’ (Thomson, 1931). Deciphering the nature of this ‘immunity’ or ‘resistance’ required more precise information about how it developed, and how it might affect ‘virulence’, drug response and transmission. Such studies might explain pronounced differences in the initial responses of malariatherapy patients: ‘the patients of our series were not of a homogeneous susceptibility to the strains . . . some clearly possessed a pristine susceptibility, others exhibited evidence of previous autochthonous infection, varying from partial immunity to complete refractoriness, the latter interpreted as an immunity homologous to the strain we employed’ (Boyd, 1942b). Similarly, the results might help to explain differences in responses to re-infections, in that ‘the usual criterion of adequate malariatherapy is the actual number of paroxysms reaching a certain febrile height . . . [so]

Strain Theory of Malaria: The First 50 Years

17

heterologous immunity exists when reinfection of a patient with a strain of P. vivax differing in geographic origin from the original strain results in definite clinical activity sufficient for the completion of a course of malaria therapy’ (Kaplan et al., 1946). Many of the early re-infection studies led to straightforward conclusions: ‘re-inoculation of a patient who has recovered from a P. falciparum attack with the same strain of the parasite will not result in a second clinical attack . . . Re-inoculation of a patient who has recovered from a P. falciparum attack with a strain of the parasite different from that which caused the primary attack, will result in a second clinical attack which may be as severe as the first’ (Boyd et al., 1936). Thus, homologous and heterologous responses were seen to reveal underlying similarities and dissimilarities between strains: a patient ‘exhibits a resistance when reinoculated homologously with the same line of parasites, but if reinoculated with parasites of the same species from a different source and presumably of a different (heterologous) line, he acquires an infection. From this it is inferred that the parasites of the second inoculation represent a different race or strain’ (Brumpt, 1949). Accordingly, crossinoculation became a common means of strain differentiation: ‘the derivation of two lines of parasites from presumably unrelated patients cannot be taken as a basis for their characterization as different strains . . . [especially] when applied to lines of parasites of obviously local indigenous origin. Their antigenic dissimilarity must be proven by the cross inoculation of convalescents’ (Boyd, 1940a). Cross-inoculation studies determined not only that ‘the White, Wilson Dam, and Mayo strains are immunologically distinct from the McCoy strain . . . [and] the Cuban, Mexican, and Panamanian strains are different from the Long strain’ (Boyd, 1940a), for instance, but, by comparing responses of patients from different regions of the United States, that ‘West Florida and contiguous Alabama are indicated as the indigenous habitat of the McCoy strain’ (Boyd and Kitchen, 1936b). Further afield, ‘little heterologous immunity was shown between vivax infections from the South Pacific, China-Burma-India theatres and from the United States. In fact, one infection originating from New Guinea exhibited little immunity to another infection from the same area, suggesting multiplicity of strains in small areas’ (Young et al., 1949). Other studies introduced complications, however. Some patients still developed fever at a fourth and parasitemia at a tenth infection with a homologous strain of P. falciparum (Ciuca et al., 1934). One possible explanation was that ‘immunity to a malarial infection exists when the subject is cured and then challenged to a reinfection with the homologous organism . . . this residual humoral immunity, however, is rapidly lost . . . [with] no evidence of cross immunity with a heterologous strain . . . [so] it is more than apparent that there is little reason to hope for an effective

18

F. Ellis McKenzie et al.

vaccine for malaria’ (Yount and Coggeshall, 1949). However, if this ‘rapid loss’ existed, the rate of loss was uncertain—for example, ‘homologous immunity to superinfection persists for as long as a year and may last seven years’ (Boyd, 1949)—and might differ with different methods of parasite transfer, for example, as ‘a temporary immunity to reinfection by the homologous organism, lasting up to two months with trophozoite inoculation and up to 13½ months with sporozoite inoculation’ (Schwartz et al., 1950). Furthermore, it began to appear that ‘convalescents from artificially induced falciparum infections usually exhibit a distinct clinical tolerance when artificially reinoculated with a heterologous strain of this parasite, manifested by a shortened period of clinical activity . . . [thus] the immunity developed during convalescence from a falciparum infection has an appreciable heterologous value’ (Boyd and Kitchen, 1945). Thus, residents of malaria-endemic countries might experience within a ‘period of time a series of episodes successively involving different strains. With a progressive series of inoculations, while the original clinical reaction might be characteristic of complete susceptibility, later episodes would exhibit characteristics of heterologous-strain immunity until the individual had acquired experience with all of the locally prevalent strains. Any inoculation subsequent to this period would result in a homologous-strain reaction’ (Kitchen, 1949). Another set of difficulties arose with ‘premunition’ (Sergent et al., 1924, 1925), the doctrine that ‘resistance to a malaria infection is dependent upon an existing infection, either active or subclinical’ (Yount and Coggeshall, 1949): that is, some responses to seeming re-infection might actually be responses to superinfection. One implication of latency or sub-detectable parasitemia (see below) was that, at least with respect to homologous immunity, ‘we never know when any one is cured. Our only proof is a renewed susceptibility to reinfection by the same strain of parasite’ (Hackett, 1937). Moreover, conclusions about homologous and heterologous responses with one species of Plasmodium might not be directly applicable to other species: ‘Superinfection with the same strain of P. vivax rarely, if ever, takes place [whereas] with P. falciparum, superinfection is apparently possible, but the course of the second infection is very greatly modified’ (Earle et al., 1939) and ‘there is less indication that acquired immunity to P. falciparum has heterologous value than in the case of acquired immunity to P. vivax’ (Boyd et al., 1936). Perhaps ‘the immunity acquired to the three species is of a different order in each case . . . and immunity to P. falciparum is the least complete. The alternative explanation is that there exists a greater variety of strains of P. falciparum, and that, while immunity to one strain is reaching a high titre, infection with a new strain, or with an older strain to which immunity has already largely disappeared, intervenes . . . [though] such a multiplicity of strains in one area seems improbable’ (Wilson, 1936).

Strain Theory of Malaria: The First 50 Years

19

Wilson (1936) would likely be astounded by the level of genetic diversity that is now known to occur within P. falciparum. Diversity is linked to levels of malaria endemicity, such that parasites from unrelated patients in Africa or New Guinea are likely to represent different strains whereas parasites from unrelated patients in South America have a higher chance of belonging to the same strain. Species-specific and strain-specific immunity is believed (although far from being formally demonstrated) to be due to (allelic) polymorphism of genes encoding the major surface proteins of each stage of the parasite, such that antibodies raised to one form of the protein bind less efficiently to heterologous forms present in other parasite species or strains. That allelic diversity is greater among P. falciparum compared to other human malarias, predicted by Wilson (1936), is being confirmed by molecular genotyping but only partially explains the lack of homologous immunity. The ability of a single strain of P. falciparum to cause numerous bouts of fever or parasitemia in a single patient (Ciuca et al., 1934) and the shorter duration of immunity following trophozoite inoculation compared with sporozoite inoculation (Schwartz et al., 1950), are likely explained by clonal antigenic variation (Dzikowski et al., 2006; Kraemer and Smith, 2006). Sequential expression among genetically identical blood stage parasites of heterologous var genes that encode for PfEMP1 expressed on the surface of infected red blood cells may enable a single parasite strain to escape effects of the developing immune response; strains with inherently high switching rates (allowing rapid sequential expression of novel antigens) would be expected to be able to cause repeated infections in the same patient. The gradual reduction in severity of disease with successive reinfections by the same strain might be explained by increasing immunity to non-variant antigens or gradual exhaustion of the repertoire of clonally variant antigens.

6.2. Clinical and parasitological response Given this accumulation of studies suggesting that ‘homologous’ protection was not necessarily immediate, absolute or permanent, that ‘heterologous’ responses might include partial protection, and so forth, more nuanced interpretations began to emerge, for example, that ‘a considerable degree of resistance to reinfection and superinfection with a homologous strain of malaria parasite, may be acquired after a single infection with such a plasmodium, and that, in some instances at least, this resistance may be increased by successive inoculations with the same strain of parasite’ (Sinton, 1940). Homologous resistance was increasingly

20

F. Ellis McKenzie et al.

described in terms of clinical severity and duration: ‘If a patient has an acute attack with a certain strain prevalent in one area and recovers from it, it is found that inoculation again with that strain shows an immunity. The attack is not so acute and he recovers more rapidly. Now if he is inoculated with a different strain brought from an area several hundred miles away, he has another acute attack similar to the first one. In other words he has some homologous immunity but no heterologous immunity’ (Bispham, 1944). By the late 1940s, it appeared that cross-inoculation might not always distinguish between heterologous strains, at least not unambiguously, without careful elaboration of the terms of reference: ‘reinoculation . . . after the chronic infection is no longer microscopically patent, with the same species and strain involved in the previous infection, is not likely to result in more than a transitory subclinical parasitemia. . . . (Boyd, 1949) Successive reinfections . . . with the same strain . . . result in progressively milder infections than the initial infection, i.e. they progressively enhance the homologous immunity . . . [and] a number of races or strains . . . may be immunologically similar. Others may not only vary in virulence and respond differently to treatment, but also fail to protect against another strain . . . An immunity may exist between strains of the same species although to a less extent than against the homologous strain . . . partial heterologous immunity between some strains and not between other strains in man has been substantiated’. One critical refinement in the terms of reference built on the observation that there were two major kinds of effects of this ‘immunity’, providing ‘some protective mechanism against both the multiplication of the parasites and their pathogenic effects. The rate of development, the efficacy, and the duration of this ‘‘immunity’’ appear to vary with (a) the species or strain of Plasmodium responsible for the infection and (b) the degree of resistance or susceptibility possessed at the time by the host’ (Sinton, 1939). There seemed to be an incomplete correspondence between these effects on symptoms and parasitemia in re-infection: ‘upon recovery from an attack of malaria . . . the patient possesses a very potent homologous immunity to that strain . . . [which] manifests itself by two characteristics: (a) the acquirement of a tolerance to densities of the parasite that in a susceptible person would produce a clinical reaction, and (b) the acquirement by the body of an ability to destroy and remove the parasites. As immunity becomes established the former characteristic is first acquired; the latter develops more slowly’ (Boyd et al., 1938b). Thus, responses to homologous and heterologous re-infection were seen as more sharply distinct in clinical than in parasitological terms, in both magnitude and duration: with ‘the reaction of ‘‘heterologous (strain) immunity’’ . . . the latent state . . . is attained as a rule in 3 to 7 days’ after the peak parasitemia, and the attacks ‘are relatively brief, usually

Strain Theory of Malaria: The First 50 Years

21

not exceeding 14–16 days in duration of clinical activity’; with ‘the ‘‘homologous (strain) immunity’’ reaction . . . the infection provokes no febrile reaction . . . [and] parasitemia is usually brief’ (Kitchen 1949). If patients could acquire ‘an immunity to fever but not to parasites’ (James and Ciuca, 1938), there might be ‘two independent kinds of immunity, one to the parasite itself, leading to its corporal destruction, and one neutralizing the pathological effects of its growth and activity upon the host’ (Hackett, 1937). In the 1930s, as the more complicated results from re-infection studies were beginning to suggest new interpretations, some authors began to use the term ‘tolerance’ specifically to refer to reduced or absent clinical response: ‘inhabitants of an area may become tolerant to the local strain of parasite, yet at the same time be susceptible to the pathogenic effects of strains present in other areas’ (Sinton, 1931). And, while the details would require investigation, ‘that the time taken to produce this tolerance is prolonged may possibly be explained if it might be assumed that there are numerous strains of parasites of malaria specifically differing in their antigenic properties’ (Thomson, 1933). Not surprisingly, premunition and other emerging issues often obscured the distinction: ‘during an attack of malaria a person acquires a tolerance to the parasite of the strain harbored, which renders him refractory to reinoculation with the same strain’ (Boyd and Stratman-Thomas, 1933a), but, while ‘a patient with a latent benign tertian infection [P. vivax] does not possess a heterologous tolerance to other strains’ (Boyd and Stratman-Thomas, 1933b), ‘following superinfection by a heterologous strain . . . we have found the attack caused by the superinfection to be less severe, indicating the actual existence of some degree of heterologous tolerance’ (Boyd and Stratman-Thomas, 1933c). However, as distinctions between clinical and parasitological aspects of response were further elaborated—for example, ‘heterologous immunity to malaria is rarely strong enough to prevent infection but it may so modify a second attack that it closely resembles a relapse’ (Russell et al., 1946)—some authors began to use the term ‘immunity’ strictly to refer to anti-parasite response, and, by the late 1940s, studies of homologous and heterologous re-infections were investigating the two aspects of response in parallel, in corresponding terms: given ‘less reaction by the host to a given bulk of infection—tolerance . . . [and] an increased ability of the host to limit the bulk of infection developed— immunity . . . [we found] that the development of tolerance preceded the development of immunity and that immunity was strain specific . . . [while] tolerance was not strictly strain specific’ (Blackburn, 1948). With respect to the ‘antigenic properties’ of parasite strains reflected in these responses, ‘the only explanation would seem to be that each of these strains contains immunological elements in common with each of the

22

F. Ellis McKenzie et al.

others’ (Hackett, 1937) and ‘that the antigenic constitution of the malaria parasites is analogous to that of some pathogenic bacteria . . . [i.e.] in all specimens of a particular strain of a malarial parasite there are parasites of two types, one containing specific antigen, the other group antigens . . . [and] that the antigens may be subdivided into 1o group antigens which are common to all members of the group, 2o group antigens which are common only to certain members of the group, 3o specific antigens peculiar to each type’ (James and Ciuca, 1938). Thus, in an attempt to integrate the evidence, the observation that ‘differences between the results of homologous and heterologous reinoculations lie mainly in the number of febrile infections produced . . . suggests that possibly the common antigenic factor in these strains may be related more to the antiparasitic element than to the anti-toxic one . . . [while the] more rapid decline of the percentage of cases showing febrile reactions after successive reinoculations than of the recorded parasitic infections . . . might suggest that the antitoxic immunity factor was developed more rapidly than the anti-parasitic one’ (Sinton, 1940). Then, ‘in the condition of premunition, while both defensive factors are in operation, the antitoxic element is probably the more efficient . . . [but] the duration of the efficacy of the anti-parasitic element persists for a longer time than does the anti-toxic one’ (Sinton, 1940). As with ‘virulence’, and drug response, several authors asked whether antigenic features of a strain might change with time. Most concluded that, since observations ‘do not suggest that extensive sexual reproduction has altered the antigenic composition of one strain . . . the antigenic characteristics of the parasites upon which immunological differentiation of strains is based, are evidently firmly fixed and retained through an indefinite number of passages through the definitive and intermediate hosts’ (Boyd, 1940a). Huff, however, argued that with ‘parasites which are distinguishable only on immunological grounds . . . whether they constitute separate races, possibly incapable of cross-breeding or whether they are simply manifestations of variations within a variety or species due to sexual reproduction can only be guessed at the present time . . . Individuals in endemic areas probably build up an immunity to one strain of malaria only to suffer from another immunologically different strain. And since there is the possibility that strains of malaria may change genetically in immunological characteristics, man in these areas is being subjected to reinfection by a multiplicity of genetic stocks of parasites. If, on the other hand, these strains of malaria have evolved far enough that they no longer cross breed it ought to be possible for individuals living in a given region to develop eventually an immunity to superinfection to all of the strains, providing the number of these strains is not unreasonably large’ (Huff, 1938).

Strain Theory of Malaria: The First 50 Years

23

The molecular basis for the two types of anti-malarial immunity— immunity against the parasite itself and tolerance to the pathology it causes—is still somewhat unclear, although the distinction is still widely appreciated (Schofield and Grau, 2005) and the former is still regarded to be somewhat more strain-specific but longer-lasting than the latter. One theory, that tolerance is provided by short-lived, T-cellindependent antibodies to non-polymorphic glycophospholipid ‘toxins’, is consistent with these observations as is the notion that immunity to parasites is provided by longer-lived and boostable antibodies to polymorphic protein antigens. Of these, conserved or relatively conserved proteins such as the circumsporozoite protein would fit the James and Ciuca (1938) definition of 1o group antigens, proteins such as merozoite surface protein 1 (MSP-1) and MSP-2 which exist as a small number of allelic families defined by conserved family-specific sequences would meet the definition of 2o group antigens, whilst the individual variants within the MSP-1 and MSP-2 families—or indeed the clonally variant PfEMP-1 proteins—could be classified as 3o group antigens. Whilst diversity within the MSP-1 and MSP-2 families seems to have arisen by a combination of both point mutation and intragenic recombination (Ferreira and Hartl, 2007; Ferreira et al., 2003), thus refuting the notion of Boyd (1940a) that antigenic characteristics are firmly fixed, it does still seem to be the case that the antigenic properties of a strain do appear to be rather more durable than—for example—drug sensitivity, suggesting perhaps that the selective forces imposed by the immune system are rather less than those imposed by chemotherapy. Regardless of the mechanism, mathematical models of epidemiological data support the view that immunity to clinical disease [tolerance] develops earlier in life than does anti-parasite immunity (Filipe et al., 2007) and that immunity to severe malaria is acquired quite rapidly (Gupta et al., 1999).

6.3. Superinfection Thus, again, there arose the question of distinguishing between re-infection and superinfection, now complicated by questions about the stability of antigens that might give rise to immunity. Using the available clinical, parasitological and immunological methods, it would be ‘necessary to know whether the presence of one parasite influences the course of infection with the others. Further, we must know if superinfection can take place and if so, at what stage of the previous infection this is possible. Finally we will need to know how many strains of each parasite there are . . .

24

F. Ellis McKenzie et al.

In nature either the number of strains or the times that they will infect in the presence of other parasites must be limited. Otherwise in endemic areas adults would have difficulty in developing immunity to all strains that existed’ (Earle et al., 1939). Sequential, single-strain infections might be just one among many sets of possibilities to be considered in interpreting responses, whatever the number of strains present. Re-infections following concurrent infections with multiple strains suggested that ‘the homologous immunity to either of two strains of P. vivax which follows simultaneous inoculation with the two strains is not so effective as that acquired after inoculation by a single strain. The immunity is characterized by the heterologous rather than the homologous properties . . . [i.e.] the simultaneous presence of two strains delays the development of an adequate homologous immunity to either’. (Boyd et al., 1938b) In World War II, it was ‘recognized that Pacific vivax malaria represents a complex situation in that each patient frequently harbors multiple strains which are not synchronized . . . the clinical features of this group of cases should be interpreted as the characteristics of multiple strain infections and possibly superposed infections of homologous strains . . . the complexity of the presence of multiple strains is such that the features of this study are only significant in a clinical light, and cannot be applied to the duration or study of immunity of a single strain infection’ (Hill and Amatuzio, 1949). The concept of multiple concurrent, superinfecting, interacting strains had implications for epidemiology and control as well as for clinical and immunological understanding. For an individual, the presence of ‘an unknown number of strains without any cross-immunity to speak of among them [implied that] only by chance is he infected twice in succession with the same kind of parasite, and his individual malaria season draws to a close only when he has solidified anew his resistance to the principal strains which are in the air around him every night. We have no idea how many strains of each parasite there are in any one locality . . . [but] any reduction in anopheline density begins by removing layer after layer of these superimposed infections before it cuts down the amount of malaria, or number of infected persons’ (Hackett, 1937). Because ‘each new strain, like a new species, finds the host defenceless and initiates a train of events culminating in an acute attack, and a period of gametocyte production . . . chronic malaria, then, is due to overlapping infections of different species and heterologous strains of plasmodia. Mixed infections must be the rule and not the exception in localities with even a moderate transmission rate . . . [and thus] children can not grow up in a malarious locality of even moderate endemicity without acquiring a representative assortment of all the species and many of the strains of plasmodia with which the local anopheles are infected . . . The transmission rate, thus determined, creates a corresponding tolerance which is made up of highly

Strain Theory of Malaria: The First 50 Years

25

specific reactions to the numerous and immunologically independent strains and species of the parasite. The picture is one of growing multiplicity of mixed infections, resulting in chronic malaria and the attendant phenomenon of mutual parasite antagonisms. The mass effect, however, of group immunity at high levels of intensity is one of powerful protection of the older age groups and the shifting of the struggle to childhood or even infancy’ (Hackett, 1941). Rates at which protective responses were acquired with age were uncertain, however, even in hyperendemic regions, as were the rates at which those responses were lost (if they were): ‘attacks are less and less frequent until the age of twelve, when, if the child survives, an immunity is produced which lasts the entire life of the native if he does not leave the locality in which he was reared. Leaving the area per se does not lessen the child’s immunity, but he would be subject to attacks of other strains of the parasite’ (Bispham, 1944) or, ‘because of the variety of strains and species, a more or less stable immunity may develop somewhere between 20 and 30 years’ (Boyd, 1949). However, given that ‘the greater the number of bites the greater the chance of the introduction of multiple strains of parasite, [which] might account for the presence of both the more virulent and the more cureresistant features recorded’ (Sinton, 1931), and that many protective responses appeared to be strain-specific, strain introductions were invoked to ‘explain the numerous observations of sudden, severe, and acute outbreaks of malaria, which can be called ‘‘malaria epidemics,’’ in endemic districts’ (Nocht and Mayer, 1937). In more general terms, the diversity of strains, clinical and parasitological effects, and frequencies of superinfection and reinfection were all seen to vary with the prevalence of infection and intensity of transmission: ‘a single or at most very few strains of parasites would be prevalent where endemicity is at a low level. The acquired immunity will render further clinical activity improbable in the event of homologous re-inoculation . . . At higher levels of endemicity the number of strains prevalent may be expected to be greater [and] through repeated reinoculations in the course of time with other strains, the individual’s immunity will become polyvalent to all of the species and strains of parasites which are locally prevalent . . . A person may be reinoculated one or more times with the same or a different species or strain of parasite, after an interval which should be expected to vary with the endemic or epidemic level prevailing . . . [if] with the same species and strain of parasite which caused a previous infection, it may result in a superinfection if homologous immunity is as yet incomplete. If effected with a different species and strain, it is distinguished as a reinfection . . . Where endemicity is at low or moderate levels, it is likely that reinoculations would be infrequent within a short interval, so that the first infection would have become latent, or even have been

26

F. Ellis McKenzie et al.

eradicated, before reinoculation. . . . Under conditions of high or hyperendemicity, or during epidemics, reinoculation may be experienced at shorter intervals . . . [which] produces a protracted or continuous infection and clinical activity’ (Boyd, 1949). While the number of strains at a given site would be expected to correlate with basic entomologic and epidemiologic variables, the same strain might be present at different sites: ‘The immunity of persons living in regions where malaria is widespread is due to the fact that ever since birth they have been inoculated by an unpredictable number of strains of the different plasmodia, some of these strains having the same antigenic properties as those which they might contract in other endemic regions, even far removed from their original surroundings’ (Brumpt, 1949). Cross-inoculation experiments in Liberia produced equivocal results: ‘we are unable therefore, to support the hypothesis that a number of immunologically differing strains exist in relatively small areas and its commonly held practical consequence that a semi-immune who travels relatively short distances in Africa is particularly liable to a ‘‘foreign’’ malaria infection with symptoms’ (Bray et al., 1962). Resistance to a seemingly novel strain might reflect some common antigenic property, or, perhaps, some innate or acquired host factors not yet identified: ‘whether the African volunteers were infected with an African strain or a Malayan strain there was no obvious difference in the symptoms or course of the malaria . . . [our results] all suggest that if there has been intensive infection throughout childhood a very definite immunity is built up which extends to a strain of P. falciparum from several hundred miles away or even from another continent . . .. the different response to infection appeared to be much more dependent on the origin, and on the quantitative degree of immunity of the patient, than on the origin of the strains’ (Davey and Robertson, 1957). Summarizing the evidence in the mid-1950s, Macdonald agreed that ‘in nature there are probably a number of strains and species of human parasite transmitted at any one moment’, but, at least in children in hyperendemic regions, ‘the occurrence of one infection with falciparum makes no material alteration to the probability of another during its course, and that fresh infections during this time are marked by a fresh onset of parasitemia materially unaffected by the previous one’ (Macdonald, 1957). He noted that responses differed between Plasmodium species, and that responses to different strains were difficult to distinguish in nature: ‘infection with P. vivax confers a homologous immunity preventing superinfection or subsequent reinfection with parasites of the same strain though not necessarily with other strains of the same species. A small degree of heterologous immunity to other strains is produced and repeated infection with several strains may ultimately produce a firm heterologous immunity to all. In the case of P. falciparum the general

Strain Theory of Malaria: The First 50 Years

27

picture is somewhat similar, but the degree of immunity conferred is considerably less. The clear-cut effects of infection with particular strains cannot be observed in the field, but only the general effect of inoculation with parasites of two or more species and of an unknown number of strains of each. In such circumstances there can be little doubt that superinfection, that is the imposition of a second infection on a first before it has died out, commonly occurs’ (Macdonald, 1957). Soon after, Bruce-Chwatt wrote that ‘it is surprising, however, that apparent superinfection should be so easy in African adults who, having been exposed to malaria since childhood could presumably have a considerable ‘‘multi-strain’’ resistance to infection . . . [so] until we have the immunological means of distinguishing between two strains of the same species of Plasmodium the explanation of such happenings and their follow-up will be difficult’ (Bruce-Chwatt, 1963). Hence, even amidst the technical advances of the 1960s, ‘it is unwise to say more than that residual immunity and not premonition [premunition] follows recovery from some infections, that the immunity is strain specific, and lasts for at least 3 years’ (Garnham, 1966). The challenge remained that ‘immunological differences exist . . . among geographically isolated strains of the same plasmodial species; but the extensive available evidence for this is based on the degree of cross-immunity among species and strains rather than on clear-cut antigenic definition’ (Zuckerman, 1964), and thus it was ‘further hoped that by serological means it will be possible to say which malaria parasites an individual has been infected with in the past, which ones he still carries, and to which strains or species he is immune’ (Voller, 1964). Though experimental infections with rodent malarias indicate withinhost competition between parasite clones (Mackinnon et al., 2002), such that one drops to sub-detectable levels when the other is introduced, no data are available on the quality of the immune response induced in such situations. Longitudinal studies in humans suggest that competition and immunity mainly affect parasite density, not the time to clearance (Sama et al., 2006); apparent loss of particular genotypes is common, as densities drop below detection limits. Molecular genotyping has confirmed the expectations that ‘mixed infections must be the rule . . . in localities with even moderate transmission rate’ (Babiker et al., 2000; Hackett, 1941; Peyerl-Hoffmann et al., 2001; Sallenave-Sales et al., 2000), that superinfection occurs more commonly in areas of higher transmission (Arnot, 1998), that novel genotypes infecting otherwise immune children give rise to symptomatic infections (Contamin et al., 1996), and that introductions of new phenotypes can cause epidemics of clinical malaria (Arez et al., 1999; Laserson et al., 1999). There is less evidence from these studies of small-scale variations in genotype

28

F. Ellis McKenzie et al.

frequencies that might lead to exposure to new strains after travelling ‘relatively short distances in Africa’ (Bray et al., 1962), but such variations have been observed in remote South American villages (Machado et al., 2004), suggesting that human population movements can create mosaics of local parasite diversity at various spatial scales. Surprisingly, despite the characterization of vast numbers of antigenic variants, the use of serological methods to determine patterns of prior exposure to different parasite strains is still in its infancy (Gray et al., 2007).

7. LATENCY AND RELAPSE With the development of malariatherapy, blood samples were sometimes taken before the onset of fever as well as after, and from these it appeared that parasites could usually be detected before the first fever occurred. Clinicians and researchers remained more focused on the incubation period (the interval between parasite inoculation and first fever in the host) than the pre-patent period (the interval between parasite inoculation and first detected parasitemia in the host blood), but noted considerable variation in the duration of each, even more with P. vivax than P. falciparum. Because the time of inoculation was known, it seemed likely that during pre-patency the parasites were not only present but multiplying at sub-detectable levels, and that the lag to the first fever involved a ‘pyrogenic threshold’. Furthermore, these intervals and levels might reflect properties of strains, arising, for instance, because ‘the number of merozoites formed at schizogony varies with these different strains of P. vivax’—referring to an average 17–18 observed with Madagascar, 16 with McCoy, 12–13 with Dutch (Boyd, 1941). Instances in which the initial latency was protracted for months were particularly striking: thus, based on ‘those very common benign tertian [P. vivax] infections of northern Europe in which even the primary attack is suppressed . . . this primary latency or prolonged incubation seems a character which belongs particularly to certain strains’ (Hackett, 1937). The pronounced differences between P. vivax strains in average incubation period—for example, 13.5 days with Madagascar and McCoy, 16.5 days with St. Elizabeth, 21 days with Dutch, 282 days with Roumanian—typically reflected differing proportions of patients with this protracted initial latency (Boyd, 1941, 1949; Kitchen, 1949). In contrast, there seemed to be relatively little variation between P. falciparum strains in average incubation period—for example, 12 days with Roman, Sardinian and West African strains, 13 days with Coker, Costa and Long strains—despite sometimes wide ranges within strains, for example, 6–25 days with Coker (Boyd, 1941; Coatney and Young, 1941; Kitchen, 1949).

Strain Theory of Malaria: The First 50 Years

29

There were complications in interpreting strain differences in incubation period, however, for instance that ‘many more sporozoites of P. vivax than of P. falciparum are required to ensure the onset of the malarial attack within the usual incubation period’ (James and Ciuca, 1938), and that ‘the duration of the incubation period tends to vary inversely with the dose of sporozoites received [and] the duration of the clinical attack appears to vary directly with the dose of sporozoites’ (Boyd, 1940b). At Horton Hospital, it appeared that though ‘latency in BT [P. vivax] malaria . . . occurs more frequently with temperate region strains than it does with tropical strains . . . true latency occurs only when the sporozoites injected are too few to set up an immediate attack’; however, with patients seemingly recovered from Madagascar-strain infections, ‘if they were infected with a different strain of the same species, some developed fever and parasites but with a protracted incubation period, usually of several months duration’ (Shute, 1946). Initial latency might be influenced not only by dose, and parasite strain, but individual host response: ‘the time which elapses between the date of being bitten by an infected mosquito and the date when the earliest clinical symptoms are felt by the patient varies with the dose of sporozoites injected, the virulence of the particular strain of parasite, and the different factors which tend to lessen or to increase the patient’s resisting power’ (James, 1920). Thus it might be, in contrast to those who ‘suggest that there is not a constant pyrogenic threshold for all strains . . . that varying susceptibility of patients rather than varying virulence of different strains of parasites is chiefly responsible for the variations in density noted at the onset’ (Boyd, 1941), and thereby the variations in incubation and pre-patent periods. When re-infection or superinfection could be excluded as possibilities, it became difficult to interpret the renewed fevers and parasitemia that sometimes followed an initial attack after latent periods of varying lengths. Several types of renewed activity could be differentiated in accord with their timing: ‘for our own purposes, and quite arbitrarily, we distinguish between the returns of fever and parasites which may follow recovery from a primary attack, thus: Recrudescence . . . Relapse . . . Recurrence’ (James, 1931), here with the first category applied to renewed activity within 8 weeks after recovery, the second 8–24 weeks after, the third more than 24 weeks after. It gradually emerged that the last of these categories—now considered ‘relapse’—might occur with P. vivax, but not P. falciparum: ‘neither latency nor long-term relapses occur in malignant tertian malaria . . . based on a study of several geographical strains of both tropical and sub-tropical origin’ (Shute, 1946). Frequencies of relapse appeared to vary dramatically between P. vivax strains, for example, ‘we have observed renewed clinical activity after cessation of the primary attack approximately ten times more often in patients inoculated with the McCoy strain than in those inoculated with

30

F. Ellis McKenzie et al.

other strains’ (Boyd, 1940a). The frequency of protracted initial latency with the Dutch strain was five-fold higher than with the Madagascar strain, and the frequency of relapses eight-fold lower (James and Ciuca, 1938). A geographical pattern emerged: ‘strains of P. vivax that originate in tropical areas characteristically produce clinical attacks of malaria at frequent intervals throughout the year . . . [but] Korean vivax malaria exhibits a bimodal pattern of clinical activity and a period of long-term latency similar to other strains of P. vivax originating in temperate climates . . . similar in all major respects to the pattern of the St. Elizabeth strain’ (Hankey et al., 1953). Thus ‘the likelihood of secondary attacks may vary with the strain . . . [and] New World strains have exhibited a decidedly less frequent tendency to become reactivated after long intervals of quiescence than have the Old World strains’ (Boyd, 1941), a difference that may have arisen because ‘in areas where transmission by mosquitoes can occur during only a short period of each year and where the infective inoculum per patient is often small, strains of P. vivax which can hibernate for many months within the human host would be much more likely to survive until the next transmission season than would strains which relapse promptly’ (Coatney et al., 1950a). Frequencies of relapse were important in evaluating potential introductions of ‘exotic’ strains of P. vivax, so studies during World War II compared relapse rates and infectivity ‘by origin of strains’—sometimes by whether the infections had most likely been acquired in islands ‘A, B or C’—reporting, for instance, that patients were most infectious during their 6th–15th relapse, and when asymptomatic (Watson, 1945). One large US clinic reported relapses in 80% of P. vivax cases returning from the Pacific, with an average latent period between recurrences of 4.2 months, compared with relapses in 30% of cases from the Mediterranean and 2% with the US St. Elizabeth strain (Schwartz et al., 1950). An extensive series of studies conducted on P. vivax from returning soldiers reported that the average pre-patent and incubation periods were shorter, and parasitemia at first fever was higher, in Mediterranean than Pacific strains, and that ‘Mediterranean malarias had a higher gametocyte density and a higher parasite level at clinical relapse than the Pacific malarias. However, the Pacific malarias showed a higher proportion of patients relapsing after treatment and a greater relative prevalence of parasitemia . . . [and therefore] the Pacific malarias might be considered as being more virulent in man than the Mediterranean malarias’ (Young et al., 1949). Thus, as with initial latency, it had become ‘apparent that the pattern of relapse in P. vivax infections is determined by the strain of parasite, as well as by immunity, chemotherapy, and the size of the infective inoculum’ (Coatney et al., 1950a), and, again, that several of these effects might be intertwined, for instance through ‘a stage of the parasite living in fixed tissue cells which intervenes between the sporozoite and the trophozoite,

Strain Theory of Malaria: The First 50 Years

31

the development of which may either be retarded, or inhibited or narcotized by a drug’ (Boyd, 1941). It appeared that ‘the frequency, duration and severity of ‘‘relapses’’ (including recrudescences) depend on the amount of ‘‘tolerance’’ or ‘‘immunity’’ which the patient may have acquired during the primary attack. Patients who are treated with quinine very early in the primary attack acquire little or no tolerance’ (James, 1931). One set of studies showed that ‘after a single attack of Chesson strain vivax malaria, cured with pentaquine–quinine, there was no appreciable homologous strain immunity. After four to seven attacks, followed by chemotherapeutic cure, homologous strain immunity was present, but was inadequate to prevent an abortive attack following sporozoite inoculation . . . [the data] strongly suggest that a large proportion of vivax infections resulting from small numbers of sporozoites will display short courses and few relapses and that they may subside under non-curative therapy without the development of homologous strain immunity’ (Coatney et al., 1950b). Another study with the Chesson strain confirmed that ‘immunological phenomena have a profound effect on the intervals between attacks and on the frequency of relapses . . . the immune cases reinoculated with the homologous parasite . . . almost invariably experienced some symptoms early . . . and also almost invariably experienced one parasite recurrence, always asymptomatic . . . relapses occurred later and less frequently after treatment of extended clinical attacks than was the case in those attacks terminated prior to extensive clinical malaria’ (Jeffery, 1956). Furthermore, ‘an important factor in relapsing malaria may be multiple mosquito bites involving perhaps a diversity of strains of the parasite . . . [and] the greater the infection dosage the greater the liability to relapse and to do so for a longer period’ (Russell et al., 1946). If so, perhaps, ‘every second, third or fourth infected Anopheles mosquito bite which is prevented means the avoidance of one, two or three relapses later on’ (Horing, 1947), with different strains. An attempt to integrate the evidence hypothesized that ‘when pre-erythrocytic parasites are discharged from an exo-erythrocytic depot, there may be several strains amongst them but one strain predominates . . . [and] immunity develops against the predominating strain, but, since there will be some overlapping of the antigenic patterns, there may also be some, possibly transient, cross-immunity against other strains. When later there is another discharge of mixed strains from the depot, the immunity that has developed against the first strain will prevent the development of the erythrocytic schizogony cycle by that strain, and another strain will predominate; this goes on until immunity has developed against all the strains present . . . The concept of multiple strains being in some way responsible for relapses is gaining ground [though] the therapeutic strains hitherto in use were probably mixed (multiple) strains; that it would be possible to initiate an infection from a single sporozoite seems improbable, but it should be possible by a

32

F. Ellis McKenzie et al.

process of dilution to induce a single-strain infection which would, if the above hypothesis is correct, be a nonrelapsing one’ (Napier, 1947). During and just after World War II, studies began using the clinical, parasitological and immunological methods available to investigate these hypotheses about connections between relapses and concurrent multiple strain infections, initially by comparing cross-inoculation responses to parasites (A, B, C, D) taken from different relapses in the same soldier. Given that ‘different strains vary characteristically in the frequency with which secondary episodes occur’ and that the effects of ‘cross inoculations with a heterologous strain’ were recognizable, ‘the results suggest that ‘‘Strains’’ C and D are closely similar antigenically, if not identical, but that ‘‘Strain’’ A substantially differs from ‘‘Strains’’ C and D’ (Boyd and Kitchen, 1948). Another study, having noted that ‘successive attacks of vivax malaria in an individual exposed in a malarious area need not necessarily be relapses caused by the same strain of parasite [and that] it is theoretically possible for the fixed-tissue parasites of two or more strains of Plasmodium vivax to coexist in a host and produce malarial attacks independently’, took advantage of ‘the characteristically different relapse patterns of the Chesson and St. Elizabeth strains of P. vivax . . . [i.e.] the Chesson strain usually produces an infection with several closely-spaced attacks, whereas the St. Elizabeth strain infection exhibits an early primary attack, several months of latency and a series of late attacks in close succession . . . Volunteers infected with the Chesson and St. Elizabeth strains of P. vivax displayed a relapse pattern consistent with a combination of the relapse patterns which were exhibited by the two strains when present separately’ (Cooper et al., 1950). Thus, with relapse as with re-infection and superinfection, it seemed reasonable that ‘if several strains were present more attacks of malaria would be required before tolerance and immunity to all strains would develop’ (Cooper et al., 1950), and, accordingly, ‘when many strains are superimposed, in individuals of differing ages and states of nutrition, who are subject to repeated reinfections as occurs in malaria in its natural habitat, there is little wonder that the pattern of relapsing vivax malaria can appear to be hopelessly complex’ (Coatney et al., 1950b). It is disconcerting to realize how recently the hypnozoite stage of P. vivax was discovered (see below) and how little progress has been made since then in understanding how hypnozoite formation and reactivation is controlled; the inability to culture P. vivax in vitro has severely limited opportunities for molecular research on this parasite. Nevertheless, the ability to genotype relapsing infections has led to confirmation that relapses show a strong tendency to be clonal and that multiple relapses in a single patient reflect reactivation of different

Strain Theory of Malaria: The First 50 Years

33

parasite genotypes (Chen et al., 2007; Imwong et al., 2007), in line with the predictions noted above. The notion that the time between being bitten by an infectious mosquito and the onset of either parasitemia or clinical symptoms is dependent, amongst others, on the number of sporozoites inoculated (James, 1920) has been supported by molecular techniques allowing accurate quantification of subpatent parasitemia (Bejon et al., 2005).

8. SUMMARY AND DISCUSSION ‘‘‘The question is,’’ said Alice, ‘‘whether you can make words mean so many different things.’’ ‘‘The question is,’’ said Humpty Dumpty, ‘‘which is to be master—that’s all’’’ (Dodgson, 1872). The historical study of strain theory provides a basis for framing contemporary debate, but the context of the debate has changed. The genetic bases for resistance to chloroquine and sulphadoxine–pyramethamine are now reasonably well-understood. The questions of infectivity and relapse remain underinvestigated. The goal of distinguishing strains in terms of clinical virulence and immunity remains as elusive as ever, in part because it is now possible to see more genetic variability than is expressed in the phenotype. The genes for merozoite stage proteins are known to be highly polymorphic, for example, but much of the observed genetic variation is neutral, or nearly so, such that multiple genotypes have essentially the same phenotype. The clinical, immunological and epidemiological relevance of genetic variability remains poorly understood. It is now evident that each P. falciparum genotype can express 50–60 different phenotypes of the PfEMP1 protein through var gene switching, and that the genes are distributed across all of the P. falciparum genome. It is plausible that heterologous/homologous immunity to P. falciparum is explained by different msp or var gene families and their patterns of crossimmunity. Human immune response may also be heterogeneous, in that the state of clinical immunity in different humans could be conferred by immunity to different sets of immunogens. Continual sexual reassortment of the genes for merozoite-stage proteins and the var genes during meiosis in the mosquito, and strong disruptive selection provide new variation. What, if anything, is a strain? If a strain could be clearly defined in one parasite generation, it might not exist in the next, and an operational definition of a strain with respect to one trait (e.g. drug resistance) might not be coherent with respect to other traits (e.g. infectivity). Aside from the limited experience of malariatherapy and laboratory experiments, the existence of a strain may be too transient for any definition to be useful. Therefore, it is unclear whether increasingly detailed genetic identification

34

F. Ellis McKenzie et al.

of strains will ever converge with the clinical definition of strains pioneered during the 1920s–1970s. The idea that each of the species that cause human malaria consists of ‘varieties, strains or races’ emerged from the observation that malaria infections seemed to differ in severity from place to place and the inference that these differences might arise from biological differences between morphologically identical parasites. When efforts to control and improve responses to malariatherapy made it necessary to distinguish and compare the parasites in clinical use, most such ‘strains’ were named on a geographic basis. The common assumption was that, in nature, ‘in each region, malaria has a character of its own conferred upon it by the peculiarities of the local parasites. No doubt there are a multitude of strains of each species of plasmodium, differing as widely in virulence, response to treatment and tendency to relapse as though they were separate species’ (Hackett, 1937). If, worldwide, ‘each parasite has many strains . . . [but] a strain prevalent in one area is frequently not found in another several hundred miles away’ (Bispham, 1944), then ‘some strains may have a localized habitat or geographical distribution’ (Boyd, 1940a). More drastically, researchers at the Horton Hospital concluded that their three decades of studies had given evidence for several strains each of P. falciparum and P. vivax, ‘from widely separated geographical areas’, but only one each of P. malariae and P. ovale, and that ‘while it is fairly certain that there are different geographical strains of malaria parasites it is extremely difficult to detect variations sufficiently well defined to justify the conclusion that different strains can exist within a single circumscribed locality’ (Shute and Maryon, 1954). The number and geographical distribution of strains—along with their infectivity and relapse characteristics—became topics of wider practical concern with the rising potential for introductions and reintroductions of malaria during World War II. Though the introduction of a particular ‘new’ strain might be probabilistic, its persistence and spread might be constrained by the relative susceptibilities of local mosquitoes or local humans. In humans, presumably, protective responses to endemic strains developed with continued exposure. Thus, ‘among the foreign infections brought into this country, there must be many different and distinct strains. As these are propagated in nature they are added to the various strains already indigenous . . . where there is little malaria now, outbreaks due to importation of foreign strains would be fairly obvious . . . [but elsewhere] the spread of foreign malaria, except under unusual and rare circumstances, could not be detected . . . As these foreign strains are immunologically distinct, it means that no protection is gained by previous infections with native malarias . . . [so] we can expect additional strains to be added to those already present in this country’ (Young et al., 1949).

Strain Theory of Malaria: The First 50 Years

35

During this same period, concurrent multiple strain infections became accepted as common and significant. Though it was well known that multiple Plasmodium species co-existed at almost all malaria-endemic sites, and that co-infecting Plasmodium species appeared to interact in suppressing each other’s parasitemia (Boyd and Kitchen, 1937, 1938), the idea that strains might similarly co-exist, co-infect and interact took hold slowly and unevenly. Understanding strain co-infections posed complex challenges: sequential infection or superinfection might elicit homologous or heterologous responses, in clinical or parasitological terms, and might influence infectivity, relapse or other features. Thus, ‘considering the opportunities for mixed infections of an unknown number of strains, it would be remarkable if any case of malaria resembled another . . . and each of these strains in turn may not be pure but may comprise a different assortment of immunological and other elements, part of which it owns in common with other strains of the same region . . . We are only at the beginning of these studies . . . [and] it is only natural that the imagination of workers in this field has seized upon the existence of such strains to explain all kinds of obscure phenomena in malaria’ (Hackett, 1937). If ‘strains’ could not be considered independent or immutable, but collections of changeable, exchangeable elements, expressed as antigenic and other properties, how could they be defined? Thus, ‘the problem of strains within species is both interesting and important, but it is by no means easy to define what is meant by this term; indeed, we have been compelled to ask ourselves, ‘‘What is a strain?’’ . . . if there are no insurmountable barriers which would prevent the spread of the parasite either by man or by the mosquito, it is the persistence of separate strains within a locality which are so difficult to comprehend, especially in hyperendemic areas . . . [since] if each strain is to retain its individuality, it must be immune to cross breeding with other strains. If several strains of a species were present in a given locality, presumably it would not be long before a large proportion of the population would be infected with two or more strains . . . [so if] a strain retains its identity in a locality where other strains occur . . . the gametocytes of one strain must be resistant to fertilization by another strain’ (Shute and Maryon, 1954). In the late 1960s, a WHO expert committee noted that ‘a parasite strain has been defined as ‘‘a population of common stock descending from a single ancestor or derived from a single source and maintained without intermixture from other sources through a number of generations’’ [WHO, 1963]. This definition may be appropriate for experimentally selected and maintained laboratory strains, but it is too precise for the present purpose’ (WHO, 1969). With ‘parasites recovered from natural infections in the field . . . the term ‘‘strain’’ is used for a population of parasites, recovered from a source in a given geographical area, that possesses confirmed or suspected

36

F. Ellis McKenzie et al.

distinctive characteristics that may be the results of the pressure of natural selection . . . [thus] biological differences in the behaviour of P. falciparum may be apparent in (a) the susceptibility to drugs, (b) the pattern of human infection, (c) the immunological response of the host and (d) the ability to infect mosquitos’ (WHO, 1969). In addition, it was thought that three types of P. vivax strains could be distinguished on the basis of their relapse patterns (WHO, 1969). That is, while a new concern about precision was noted, and discounted, the listing of strain characteristics remained identical to that given in the early 1930s. In the 1970s and early 1980s, when it became possible to clone and cultivate P. falciparum (Rosario, 1981; Trager and Jensen, 1976; Trager et al., 1981), it became clear that virtually every natural ‘isolate’ contained a mixture of parasite entities, each of which when cloned and cultivated might demonstrate strikingly different phenotypic properties with respect to growth rate, drug susceptibility, gametocyte production, antigen and enzyme variants (Burkot et al., 1984; Graves et al., 1984b). This wide range of ‘clones’ was seen to reflect ‘the extent of the genetic diversity which can exist within a single isolate, or ‘‘strain,’’ of these parasites’; thus, were an isolate cultured in conditions which favored the growth of some clones over others, ‘some ‘‘strains’’ of P. falciparum might well undergo changes in such characters as drug resistance or antigenic phenotype’ (Thaithong et al., 1984). Similarly, the report of a 1981 WHO expert committee on ‘malaria parasite strain characterization’ considered ‘isolate’ and ‘strain’ synonymous, noting that ‘the advent of cloning of asexual blood forms of P. falciparum is expected to provide a number of well-characterized uniform strains’ with respect to ‘the available biological markers, including drug sensitivity, isoenzymes, antigenic determinants, plasmodial infectivity to insect vectors, and DNA and other biochemical characteristics’ (WHO, 1981). If an ‘isolate’ or ‘strain’ was a community of parasite entities, however, it was not at all clear what constraints preserved its character from one host to the next. If ‘phenotype’ was a matter of proportions within a mixture, then it might be the uniformity or diversity of hosts and transmission between them that determined phenotypic stability by shaping those proportions: hosts were mixing vessels, sampled by mosquitoes. But, for parasites with an obligate sexual phase in the mosquito, it was not clear how ‘clones’ could be the constituent entities in nature. The questions have since shifted to (multi-locus) genotypes, but they still echo aspects of strain theory from long ago, for example, that ‘partial immunities, partial tolerances might be explained by the loss of certain elements during the passages of composite strains’ (Hackett, 1937). Much of our current knowledge of malaria derives from investigations into the ‘obscure phenomena’ of strains. This is most obviously so in

Strain Theory of Malaria: The First 50 Years

37

malaria immunology—for example, in distinguishing between homologous and heterologous, clinical and parasitological aspects of response, or in proposing that responses might differ due to premunition or inherent and acquired differences between hosts. Not until the mid-1970s, for instance, could the insusceptibility of black patients to P. vivax malariatherapy, and the rarity of P. vivax infections in Sub-Saharan Africa, be attributed to an absence of Duffy receptor (Miller et al., 1976)—a finding counter to the suggestion, noted above, that ‘the human intermediate host is not likely to be a factor in limiting the extension of the range of strains of these parasites’ (Boyd et al., 1938a). Similarly, over time, the observation that while incubation and prepatent periods might vary between strains, they were generally shorter with blood than with sporozoite inoculation—that ‘the method of inoculation (natural or artificial) appears to have some influence on subsequent events’ (Stratman-Thomas, 1941)—led to the hypothesis that the natural parasite life-cycle included an intermediate stage in fixed tissue: it ‘may be that by blood inoculation only the forms of the parasite which live in red blood corpuscles are introduced, whilst by the natural method of infection a form of the parasite is introduced which has always lived, not in red blood corpuscles, but in tissue cells . . . true (long interval) relapses, and recurrences are not observed (so far as we can ascertain) in inoculated cases, while they occur in 50 per cent of mosquito-infected cases. This difference has led to various suggestions about what happens to sporozoites when they are injected by the mosquito’ (James, 1931). Naturally-induced infections often appeared less responsive to drugs, less infectious to mosquitoes and less resistant to re-infection during the initial latent period than later (James and Ciuca, 1938), but hepatic forms of P. vivax and P. falciparum were not detected until the late 1940s (Shortt and Garnham, 1948; Shortt et al., 1951), and P. vivax hypnozoites not until the 1980s (Krotoski et al., 1982). Yet it can still be asked, as it was 50 years ago, ‘Is not the word ‘‘strain’’ in regard to malaria used much too loosely?’ (Shute, 1958). Beginning in the 1920s, the theory was that ‘strains’ exist, defined by distinct, observable characteristics. Over the next 50 years, attempts to tighten ‘loose’ definitions produced important insights into clinical virulence, infectivity, reaction to anti-malarial remedies, antigenic properties, latency and relapse—among them the insights that, upon closer examination, some characteristics that had seemed evident might prove too elusive to be useful in definition. Strain theory had seemed to make testable predictions, but the insights it produced arose from its ramifying ambiguities, and from the recognition of ever more complex confounding variables. What, if anything, will be said of the second 50 years of strain theory? The first dictionary of the English language defined ‘theory’as ‘speculation; not practice; scheme; plan or system yet subsisting only in the mind’

38

F. Ellis McKenzie et al.

(Johnson, 1755). Atoms, genes and germs were once ‘only in the mind’, with definitions that were sometimes ‘too loose’ or ‘too precise’, but those terms have persisted as the concepts were refined and explanations tested, with profound practical consequences. Other terms—for example, ‘phlogiston’ and ‘humours’—have been abandoned. If the four species that cause human malaria contain coherent units that behave as ‘strains’, interventions would likely be improved by understanding them. Progress towards that understanding requires a testable, falsifiable theory of Plasmodium ‘strains’, no less now, than in the 1920s.

ACKNOWLEDGEMENTS We gratefully acknowledge the contributions of J. Makulowich, B. C. Sorkin and M. Taylor.

REFERENCES A-Elbasit, I. E., Elghazali, G., A-Elgadir, T. M., Hamad, A. A., Babiker, H. A., Elbashir, M. I., and Giha, H. A. (2007). Allelic polymorphism of MSP2 gene in severe P. falciparum malaria in an area of low and seasonal transmission. Parasitol. Res. 102, 29–34. Anderson, T. J., and Roper, C. (2005). The origins and spread of antimalarial drug resistance: Lessons for policy makers. Acta Trop. 94, 269–280. Arez, A. P., Snounou, G., Pinto, J., Sousa, C. A., Modiano, D., Ribeiro, H., Franco, A. S., Alves, J., and do Rosario, V. E. (1999). A clonal Plasmodium falciparum population in an isolated outbreak of malaria in the Republic of Cabo Verde. Parasitology 118, 347–355. Arnot, D. (1998). Unstable malaria in Sudan: The influence of the dry season. Clone multiplicity of Plasmodium falciparum infections in individuals exposed to variable levels of disease transmission. Trans. R. Soc. Trop. Med. Hyg. 92, 580–585. Babiker, H. A., Ranford-Cartwright, L. C., Currie, D., Charlwood, J. D., Billingsley, P., Teuscher, T., and Walliker, D. (1994). Random mating in a natural population of the malaria parasite Plasmodium falciparum. Parasitology 109, 413–421. Babiker, H. A., Abdel-Muhsin, A. A., Hamad, A., Mackinnon, M. J., Hill, W. G., and Walliker, D. (2000). Population dynamics of Plasmodium falciparum in an unstable malaria area of eastern Sudan. Parasitology 120, 105–111. Badenski, G. T. (1936). ‘‘A Note Describing the Technique Employed in Malariatherapy in the Centres at Rome and Horton (England).’’ League of Nations Health Organisation Malaria Commission, Geneva(document 237). Barnes, K. I., and White, N. J. (2005). Population biology and antimalarial resistance: The transmission of antimalarial drug resistance in Plasmodium falciparum. Acta Trop. 94, 230–240. Bejon, P., Andrews, L., Andersen, R. F., Dunachie, S., Webster, D., Walther, M., Gilbert, S. C., Peto, T., and Hill, A. V. (2005). Calculation of liver-to-blood inocula, parasite growth rates, and preerythrocytic vaccine efficacy, from serial quantitative polymerase chain reaction studies of volunteers challenged with malaria sporozoites. J. Infect. Dis. 191, 619–626. Bhasin, V. K., and Trager, W. (1984). Gametocyte-forming and non-gametocyte-forming clones of Plasmodium falciparum. Am. J. Trop. Med. Hyg. 33, 534–537. Bishop, A. (1955). Problems concerned with gametogenesis in Haemosporidiidea, with particular reference to the genus Plasmodium. Parasitology 45, 163–185. Bispham, W. N. (1944). ‘‘Malaria.’’ Williams and Wilkins Company, Baltimore.

Strain Theory of Malaria: The First 50 Years

39

Blackburn, C. R. B. (1948). Observations on the development of resistance to vivax malaria. Trans. R. Soc. Trop. Med. Hyg. 42, 117–162. Boyd, M. F. (1940a). On strains or races of the malaria parasites. Am. J. Trop. Med. 20, 69–80. Boyd, M. F. (1940b). The influence of sporozoite dosage in vivax malaria. Am. J. Trop. Med. 20, 279–286. Boyd, M. F. (1941). The infection in the intermediate host: Symptomatology, general considerations. In ‘‘A Symposium on Human Malaria’’ (F. R. Moulton, ed.), pp. 163–182. American Association for the Advancement of Science, Washington DC. Boyd, M. F. (1942a). On the varying infectiousness of different patients infected with vivax malaria. Am. J. Trop. Med. 22, 73–81. Boyd, M. F. (1942b). Criteria of immunity and susceptibility in naturally induced vivax malaria infections. Am. J. Trop. Med. 22, 217–226. Boyd, M. F. (1949). Epidemiology of malaria: Factors related to the intermediate host. In ‘‘Malariology’’ (M. F. Boyd, ed.), pp. 551–607. W. B. Saunders Company, Philadelphia. Boyd, M. F., and Kitchen, S. F. (1936a). The comparative susceptibility of Anopheles quadrimaculatus, Say, and Anopheles punctipennis, Say, to Plasmodium vivax, Grassi, and Plasmodium falciparum, Welch. Am. J. Trop. Med. 16, 67–71. Boyd, M. F., and Kitchen, S. F. (1936b). On the localization of the geographical distribution of the McCoy strain of Plasmodium vivax. Am. J. Trop. Med. 16, 583–587. Boyd, M. F., and Kitchen, S. F. (1937). Simultaneous inoculation with Plasmodium vivax and Plasmodium falciparum. Am. J. Trop. Med. 17, 855–861. Boyd, M. F., and Kitchen, S. F. (1938). Vernal vivax activity in persons simultaneously inoculated with Plasmodium vivax and Plasmodium falciparum. Am. J. Trop. Med. 18, 505–514. Boyd, M. F., and Kitchen, S. F. (1945). On the heterologous value of acquired immunity to Plasmodium falciparum. J. Natl. Malar. Soc. 4, 301–306. Boyd, M. F., and Kitchen, S. F. (1948). On the homogeneity or heterogeneity of Plasmodium vivax infections acquired in highly endemic regions. Am. J. Trop. Med. 28, 29–34. Boyd, M. F., and Stratman-Thomas, W. K. (1933a). Studies on benign tertian malaria 1. On the occurrence of acquired tolerance to Plasmodium vivax. Am. J. Hyg. 17, 55–59. Boyd, M. F., and Stratman-Thomas, W. K. (1933b). Studies on benign tertian malaria 3. On the absence of a heterologous tolerance to Plasmodium vivax. Am. J. Hyg. 18, 482–484. Boyd, M. F., and Stratman-Thomas, W. K. (1933c). Studies on benign tertian malaria 6. On heterologous tolerance. Am. J. Hyg. 20, 482–487. Boyd, M. F., and Stratman-Thomas, W. K. (1934). The comparative susceptibility of A. quadrimaculatus, Say, and A. crucians, Wied (inland variety) to the parasites of human malaria. Am. J. Hyg. 20, 247–257. Boyd, M. F., Stratman-Thomas, W. K., and Kitchen, S. F. (1935). On the relative susceptibility of Anopheles quadrimaculatus to Plasmodium vivax and Plasmodium falciparum. Am. J. Trop. Med. 15, 485–493. Boyd, M. F., Stratman-Thomas, W. K., and Kitchen, S. F. (1936). On acquired immunity to Plasmodium falciparum. Am. J. Trop. Med. 16, 139–145. Boyd, M. F., Carr, H. P., and Rozeboom, L. E. (1938a). On the comparative susceptibility of certain species of Nearctic and Neotropical anophelines to certain strains of P. vivax and P. falciparum from the same regions. Am. J. Trop. Med. 18, 157–168. Boyd, M. F., Kupper, W. H., and Matthews, C. B. (1938b). A deficient homologous immunity following simultaneous inoculation with two strains of Plasmodium vivax. Am. J. Trop. Med. 18, 521–524. Boyd, M. F., Kitchen, S. F., and Matthews, C. B. (1941). On the natural transmission of infection from patients concurrently infected with two strains of Plasmodium vivax. Am. J. Trop. Med. 21, 645–652.

40

F. Ellis McKenzie et al.

Bray, R. S., Gunders, A. E., Burgess, R. W., Freeman, J. B., Etzel, E., Guttuso, C., and Colussa, B. (1962). The inoculation of semi-immune African with sporozoites of Laverania falcipara (Plasmodium falciparum) in Liberia. Riv. Malariol. 41, 199–210. Bruce-Chwatt, L. J. (1963). A longitudinal survey of natural malaria infection in a group of West African adults, II. West Afr. Med. J. 12, 199–217. Brumpt, E. (1949). The human parasites of the genus Plasmodium. In ‘‘Malariology’’ (M. F. Boyd, ed.), pp. 65–121. W. B. Saunders Company, Philadelphia. Bull, P. C., Pain, A., Ndungu, F. M., Kinyanjui, S. M., Roberts, D. J., Newbold, C. I., and Marsh, K. (2005). Plasmodium falciparum antigenic variation: Relationships between in vivo selection, acquired antibody response, and disease severity. J. Infect. Dis. 192, 1119–1126. Bunker, H. A., Jr., and Kirby, G. H. (1925). Treatment of general paralysis by inoculation with malaria. J. Am. Med. Assoc. 84, 563–568. Burgess, R. W., and Young, M. D. (1950). The comparative susceptibility of Anopheles quadrimaculatus and Anopheles freeborni to infection by Plasmodium vivax (St. Elizabeth strain). J. Natl. Malar. Soc. 9, 218–221. Burkot, T. R., Williams, J. L., and Schneider, I. (1984). Infectivity to mosquitoes of Plasmodium falciparum clones grown in vitro from the same isolate. Trans. R. Soc. Trop. Med. Hyg. 78, 339–341. Celli, A. (1901). ‘‘Malaria.’’ Longmans, Green, and Co., London. Chen, N., Auliff, A., Rieckmann, K., Gatton, M., and Cheng, Q. (2007). Relapses of Plasmodium vivax infection result from clonal hypnozoites activated at predetermined intervals. J. Infect. Dis. 195, 934–941. Chernin, E. (1984). The malariatherapy of neurosyphilis. J. Parasitol. 70, 611–617. Cicero, M. T. (45BC) De Natura Deorum. In ‘‘Cicero XIX’’ (H. Rackham, transl.), part 3. Harvard University Press, Cambridge, MA, 1933. Ciuca, M., Ballif, L., and Chelarescu-Vieru, M. (1934). Immunity in malaria. Trans. R. Soc. Trop. Med. Hyg. 27, 619–622. Clyde, D. F., and Shute, G. T. (1954). Resistance of East African varieties of Plasmodium falciparum to pyrimethamine. Trans. R. Soc. Trop. Med. Hyg. 48, 495–500. Clyde, D. F., Dawkins, A. T., Jr., Heiner, G. G., McCarthy, V. C., and Hornick, R. B. (1969). Characteristics of four new drug-resistant strains of Plasmodium falciparum from South-east Asia. Mil. Med. 134, 787–794. Coatney, G. R., and Young, M. D. (1941). The taxonomy of the human malaria parasites with notes on the principal American strains. In ‘‘A Symposium on Human Malaria’’ (F. R. Moulton, ed.) pp. 19–24. American Association for the Advancement of Science, Washington DC. Coatney, G. R., Cooper, W. C., Ruhe, D. S., Young, M. D., and Burgess, R. W. (1950a). Studies in human malaria. XVIII. The life pattern of sporozoite-induced St. Elizabeth strain vivax malaria. Am. J. Hyg. 51, 200–215. Coatney, G. R., Cooper, W. C., and Young, M. D. (1950b). Studies in human malaria. XXX. A summary of 204 sporozoite-induced infections with the Chesson strain of Plasmodium vivax. J. Natl. Malar. Soc. 9, 381–396. Coatney, G. R., Collins, W. E., Warren, Mc W., and Contacos, P. G. (1971). ‘‘The Primate Malarias.’’ U. S. Department of Health Education and Welfare, Washington DC. Collins, W. E., and Jeffery, G. M. (1999). A retrospective examination of sporozoite-and trophozoite-induced infections with Plasmodium falciparum. Am. J. Trop. Med. Hyg. 61, 4–48. Collins, W. E., Jeffery, G. M., and Burgess, R. W. (1963). Comparative infectivity of strains of P. falciparum to Anopheles quadrimaculatus. Mosq. News 23, 102–104. Contacos, P. G., Lunn, J. S., and Coatney, G. R. (1963). Drug-resistant falciparum malaria from Cambodia and Malaya. Trans. R. Soc. Trop. Med. Hyg. 57, 417–424.

Strain Theory of Malaria: The First 50 Years

41

Contamin, H., Fandeur, T., Rogier, C., Bonnefoy, S., Konate, L., Trape, J. F., and MercereauPuijalon, O. (1996). Different genetic characteristics of Plasmodium falciparum isolates collected during successive clinical malaria episodes in Senegalese children. Am. J. Trop. Med. Hyg. 54, 632–643. Cooper, W. C., Coatney, G. R., Culwell, W. B., Eyles, D. E., and Young, M. D. (1950). Studies in human malaria. XXVI. J. Natl. Malar. Soc. 9, 187–194. Craig, C. F. (1926). ‘‘Parasitic Protozoa of Man.’’ J. P. Lippincott Company, Philadelphia. Davey, D. G., and Robertson, G. I. (1957). Experiments with antimalarial drugs in man II. A description of methods. Trans. R. Soc. Trop. Med. Hyg. 51, 450–456. Day, K. P., Karamalis, F., Thompson, J., Barnes, D. A., Peterson, C., Brown, H., Brown, G. V., and Kemp, D. J. (1993). Genes necessary for expression of a virulence determinant and for transmission of Plasmodium falciparum are located on a 0.3-magabase region of chromosome 9. Proc. Natl. Acad. Sci. USA 90, 8292–8296. de Rudolf, G. M. (1924). The temperature-chart in artificially-inoculated malaria. J. Trop. Med. Hyg. 27, 259–263. de Rudolf, G. M. (1927). ‘‘Therapeutic Malaria.’’ Oxford University Press, London. Deaderick, W. H. (1909). ‘‘A Practical Study of Malaria.’’ W. B. Saunders Company, Philadelphia. Dobano, C., Rogerson, S. J., Taylor, T. E., McBride, J. S., and Molyneux, M. E. (2007). Expression of merozoite surface protein markers by Plasmodium falciparum-infected erythrocytes in peripheral blood and tissues of children with fatal malaria. Infect. Immun. 75, 643–652. Dodgson, C. L. (1872). ‘‘Through the Looking-Glass.’’ Macmillan and Co, London. Dzikowski, R., Templeton, T. J., and Deitsch, K. (2006). Variant antigen gene expression in malaria. Cell. Microbiol. 8, 1371–1381. Earle, W. C., Perez, M., del Rio, J., and Arzola, C. (1939). Observations on the course of naturally acquired malaria in Puerto Rico. Puerto Rican J. Public Health Trop. Med. 14, 391–406. Edeson, J. F. B., and Field, J. W. (1950). Proguanil-resistant falciparum malaria in Malaya. Br. Med. J. ii, 147–151. Ehrman, F. C., Ellis, J. M., and Young, M. D. (1945). Plasmodium vivax Chesson strain. Science 101, 377. Fairley, N. H. (1946). Atebrin susceptibility of the Aitaipe-Wewak strains of P. falciparum and P. vivax. Trans. R. Soc. Trop. Med. Hyg. 40, 229–273. Ferreira, M. U., and Hartl, D. L. (2007). Plasmodium falciparum: Worldwide sequence diversity and evolution of the malaria vaccine candidate merozoite surface protein-2 (MSP-2). Exp. Parasitol. 115, 32–40. Ferreira, M. U., Ribeiro, W. L., Tonon, A. P., Kawamoto, F., and Rich, S. M. (2003). Sequence diversity and evolution of the malaria vaccine candidate merozoite surface protein-1 (MSP-1) of Plasmodium falciparum. Gene 304, 65–75. Fiertz, C. O. (1926). The malarial treatment of general paralysis. State Hosp. Q. xi, 626–643. Filipe, J. A. N., Riley, E. M., Drakeley, C. J., Sutherland, C. J., and Ghani1, A. C. (2007). Determination of the processes driving the acquisition of immunity to malaria using a mathematical transmission model. PLoS Comput. Biol. 3, e255. Francis, S. E., Malkov, V. A., Oleinikov, A. V., Rossnagle, E., Wendler, J. P., Mutabingwa, T. K., Fried, M., and Duffy, P. E. (2007). Six genes are preferentially transcribed by the circulating and sequestered forms of Plasmodium falciparum that infect pregnant women. Infect. Immun. 75, 4838–4850. Garnham, P. C. C. (1966). ‘‘Malaria Parasites and other Haemosporidia.’’ Blackwell Scientific Publications, Oxford. Gatton, M. L., and Cheng, Q. (2004). Modeling the development of acquired clinical immunity to Plasmodium falciparum malaria. Infect. Immun. 72, 6538–6545.

42

F. Ellis McKenzie et al.

Golgi, C. (1889). On the cycle of development of malarial parasites in tertian fever: Differential diagnosis between the intracellular malarial parasites of tertian and quartan fever. In ‘‘Tropical Medicine and Parasitology: Classic Investigations’’ (B. H. Kean, K. E. Mott, and A. J. Russell, eds.), Volume I, pp. 26–35. Cornell University Press, Ithaca, NY, 1978. Grant, A. R. (1923). The treatment of general paralysis by malaria. Br. Med. J. ii, 698–700. Graves, P. M., Carter, R., and McNeill, K. M. (1984a). Gametocyte production in cloned lines of Plasmodium falciparum. Am. J. Trop. Med. Hyg. 33, 1045–1050. Graves, P. M., Carter, R., Keystone, J. S., and Seeley, D. C., Jr. (1984b). Drug sensitivity and isoenzyme type in cloned lines of Plasmodium falciparum. Am. J. Trop. Med. Hyg. 33, 212–219. Gray, J. C., Corran, P. H., Mangia, E., Gaunt, M. W., Li, Q., Tetteh, K. K., Polley, S. D., Conway, D. J., Holder, A. A., Bacarese-Hamilton, T., Riley, E. M., and Crisanti, A. (2007). Profiling the antibody immune response against blood stage malaria vaccine candidates. Clin. Chem. 53, 1244–1253. Gregson, A., and Plowe, C. V. (2005). Mechanisms of resistance of malaria parasites to antifolates. Pharmacol. Rev. 57, 117–145. Gupta, S., and Day, K. P. (1994a). A strain theory of malaria transmission. Parasitol. Today 10, 476–481. Gupta, S., and Day, K. P. (1994b). A theoretical framework for the immunoepidemiology of Plasmodium falciparum malaria. Parasite Immunol. 16, 361–370. Gupta, S., Snow, R. W., Donnelly, C. A., Marsh, K., and Newbold, C. (1999). Immunity to non-cerebral severe malaria is acquired after one or two infections. Nat. Med. 5, 340–343. Hackett, L. W. (1937). ‘‘Malaria in Europe.’’ Oxford University Press, London. Hackett, L. W. (1941). Malaria and the community. In ‘‘A Symposium on Human Malaria’’ (F. R. Moulton, ed.), pp. 148–156. American Association for the Advancement of Science, Washington DC. Hankey, D. D., Jones, R., Jr., Coatney, G. R., Alviing, A. S., Coker, W. G., Garrison, P. L., and Donovan, W. N. (1953). Korean vivax malaria. I. Am. J. Trop. Med. Hyg. 2, 958–969. Hastings, I. M., and Donnelly, M. J. (2005). The impact of antimalarial drug resistance mutations on parasite fitness, and its implications for the evolution of resistance. Drug Resist. Updat. 8, 43–50. Hehir, P. (1927). ‘‘Malaria in India.’’ Oxford University Press, London. Hill, E., and Amatuzio, D. S. (1949). Southwest Pacific vivax malaria. Am. J. Trop. Med. 29, 203–214. Horing, R. O. (1947). Induced and war malaria. J. Trop. Med. Hyg. 50, 150–160. Huber, W., Haji, H., Charlwood, J. D., Certa, U., Walliker, D., and Tanner, M. (1998). Genetic characterization of the malaria parasite Plasmodium falciparum in the transmission from the host to the vector. Parasitology 116, 95–101. Huff, C. G. (1938). The significance of different strains of malaria and mosquitoes in the epidemiology of the disease. Am. J. Med. Technol. 4, 41–46. Imwong, M., Snounou, G., Pukrittayakamee, S., Tanomsing, N., Kim, J. R., Nandy, A., Guthmann, J. P., Nosten, F., Carlton, J., Looareesuwan, S., Nair, S., Sudimack, D., et al. (2007). Relapses of Plasmodium vivax infection usually result from activation of heterologous hypnozoites. J. Infect. Dis. 195, 927–933. James, S. P. (1920). ‘‘Malaria at Home and Abroad.’’ John Bale, Sons & Danielsson, Ltd., London. James, S. P. (1931). Some general results of a study of induced malaria in England. Trans. R. Soc. Trop. Med. Hyg. 27, 477–538. James, S. P., and Ciuca, M. (1938). Species and races of human malaria parasites and a note on immunity. In ‘‘Acta Conventus Tertii de Tropicis Atque Malariae Morbis.’’ Volume 2, pp. 269–281. C. A. Spin & Zoon N. V., Amsterdam.

Strain Theory of Malaria: The First 50 Years

43

James, S. P., Nicol, W. B., and Shute, P. G. (1932). A study of induced malignant tertian malaria. Proc. R. Soc. Med. 25, 1153–1186. James, S. P., Nicol, W. B., and Shute, P. G. (1936). Clinical and parasitological observations on induced malaria. Proc. R. Soc. Med. 29, 879–894. Jeffery, G. M. (1951). Observations on a gametocyteless strain of Plasmodium falciparum. J. Natl. Malar. Soc. 10, 337–344. Jeffery, G. M. (1956). Relapses with Chesson strain Plasmodium vivax following treatment with chloroquine. Am. J. Trop. Med. Hyg. 5, 1–13. Johnson, S. (1755). ‘‘A Dictionary of the English Language.’’ James Maxwell, London. Kaplan, L. I., Read, H. S., and Becker, F. T. (1946). Homologous and heterologous strains of Plasmodium vivax: A cross-inoculation study of malaria strain immunity. J. Lab. Clin. Med. 31, 400–408. Kenrick, W. H. (1910). Malaria in the Central Provinces. Proceedings of the Imperial Malaria Conference, pp. 22–24. Government Central Branch Press, Simla. Kitchen, S. F. (1949). Vivax malaria. In ‘‘Malariology’’ (M. F. Boyd, ed.), pp. 1027–1045. W. B. Saunders Company, Philadelphia. Kraemer, S. M., and Smith, J. D. (2006). A family affair: var genes, PfEMP1 binding, and malaria disease. Curr. Opin. Microbiol. 9, 374–380. Krotoski, W. A., Collins, W. E., Bray, R. S., Garnham, P. C., Cogswell, F. B., Gwadz, R. W., Killick-Kendrick, R., Wolf, R., Sinden, R., Koontz, L. C., and Stanfill, P. S. (1982). Demonstration of hypnozoites in sporozoite-transmitted Plasmodium vivax infection. Am. J. Trop. Med. Hyg. 31, 1291–1293. Kublin, J. G., Cortese, J. F., Njunju, E. M., Mukadam, R. A., Wirama, J. J., Kazembe, P. N., Djimde, A. A., Kouriba, B., Taylor, T. E., and Plowe, C. V. (2003). Reemergence of chloroquine-sensitive Plasmodium falciparum malaria after cessation of chloroquine use in Malawi. J. Infect. Dis. 187, 1870–1875. Kupper, W. H. (1939). ‘‘The Malarial Therapy of General Paralysis and Other Conditions.’’ Edwards Brothers, Inc., Ann Arbor MI. Kyriacou, H. M., Stone, G. N., Challis, R. J., Raza, A., Lyke, K. E., Thera, M. A., Kone, A. K., Doumbo, O. K., Plowe, C. V., and Rowe, J. A. (2006). Differential var gene transcription in Plasmodium falciparum isolates from patients with cerebral malaria compared to hyperparasitaemia. Mol. Biochem. Parasitol. 150, 211–218. Lambrechts, L., Halbert, J., Durand, P., Gouagna, L. C., and Koella, J. C. (2005). Host genotype by parasite genotype interactions underlying the resistance of anopheline mosquitoes to Plasmodium falciparum. Malar. J. 11, 3. Laserson, K. F., Petralanda, I., Almera, R., Barker, R. H., Jr., Spielman, A., Maguire, J. H., and Wirth, D. F. (1999). Genetic characterization of an epidemic of Plasmodium falciparum malaria among Yanomami Amerindians. J. Infect. Dis. 180, 2081–2085. Laveran, A. (1893). ‘‘Paludism.’’ The New Syndenham Society, London. League of Nations Health Organisation Malaria Commission (1934). Ten years of activity of the Commission Geneva (document 212). Leslie, J. W. T. (1910). Malaria in India. In ‘‘Proceedings of the Imperial Malaria Conference’’, pp. 3–11. Government Central Branch Press, Simla. Lilly, G. A. (1925). The treatment of general paralysis at Hanwell Mental Hospital. J. Ment. Sci. 71, 267–278. Lomholt, M. (1944). ‘‘Clinic and Prognosis of Malaria-Treated Paralysis.’’ Ejnar Munksgaard, Copenhagen. Macbride, H. J. (1924). The treatment of general paralysis of the insane by malaria. J. Neurol. Psychopathol. 5, 13–27. Macdonald, G. (1957). ‘‘The Epidemiology and Control of Malaria.’’ Oxford University Press, London.

44

F. Ellis McKenzie et al.

Machado, R. L., Povoa, M. M., Calvosa, V. S., Ferreira, M. U., Rossit, A. R., dos Santos, E. J., and Conway, D. J. (2004). Genetic structure of Plasmodium falciparum populations in the Brazilian Amazon region. J. Infect. Dis. 190, 1547–1555. Mackinnon, M. J., and Read, A. F. (1999). Selection for high and low virulence in the malaria parasite Plasmodium chabaudi. Proc. R. Soc. London Ser. B 266, 741–748. Mackinnon, M. J., Gaffney, D. J., and Read, A. F. (2002). Virulence in rodent malaria: Host genotype by parasite genotype interactions. Infect. Genet. Evol. 1, 287–296. Marchiafava, E., and Bignami, A. (1894). ‘‘On Summer-Autumn Malarial Fevers.’’ The New Syndenham Society, London. Marchoux, E. (1922). Multiplicite des races dans les trois formes de parasites du paludisme. Bulletin de la Socie´te´ de Pathologie Exotique 15, 108–109. McNamara, J. V., Rieckmann, K. H., Frischer, H., Stockert, T. A., Carson, P. E., and Powell, R. D. (1967). Acquired decrease in sensitivity to quinine observed during studies with a strain of chloroquine-resistant Plasmodium falciparum. Ann. Trop. Med. Parasitol. 61, 386–395. Miller, L. H., Mason, S. J., Clyde, D. F., and McGinniss, M. H. (1976). The resistance factor to Plasmodium vivax in blacks. The Duffy-blood-group genotype, FyFy. N. Engl. J. Med. 295, 302–304. Mita, T., Kaneko, A., Hwaihwanje, I., Taukahara, T., Takahashi, N., Osawa, H., Tanabe, K., Kobayakawa, T., and Bjorkman, A. (2006). Rapid selection of dhfr mutant allele in Plasmodium falciparum isolates after the introduction of sulfadoxine/pyrimethamine in combination with 4-aminoquinolones in Papua New Guinea. Infect. Genet. Evol. 6, 447–452. Moore, D. V., and Lanier, J. E. (1961). Observations on two Plasmodium falciparum infections with an abnormal response to chloroquine. Am. J. Trop. Med. Hyg. 10, 5–9. Napier, L. E. (1947). The malaria relapse problem. J. Trop. Med. Hyg. 50, 147–150. Nocht, B., and Mayer, M. (1937). ‘‘Malaria.’’ John Bale Medical Publications, London. Pearce, R., Malisa, A., Kachur, S. P., Barnes, K., Sharp, B., and Roper, C. (2005). Reduced variation around drug-resistant dhfr alleles in African Plasmodium falciparum. Mol. Biol. Evol. 22, 1834–1844. Peyerl-Hoffmann, G., Jelinek, T., Kilian, A., Kabagambe, G., Metzger, W. G., and von Sonnenburg, F. (2001). Genetic diversity of Plasmodium falciparum and its relationship to parasite density in an area with different malaria endemicities in West Uganda. Trop. Med. Int. Health 6, 607–613. Pijper, A., and Russell, E. D. (1926). Malaria treatment of general paralysis: A report on 44 cases. S. Afr. Med.l Rec. 35, 292–305. Ramsdale, C. D., and Coluzzi, M. (1975). Studies on the infectivity of tropical African strains of Plasmodium falciparum to some southern European vectors of malaria. Parassitologia 17, 39–48. Ranford-Cartwright, L. C., Balfe, P., Carter, R., and Walliker, D. (1993). Frequency of crossfertilization in the human malaria parasite Plasmodium falciparum. Parasitology 107, 11–18. Robertson, G. I., Davey, D. G., and Fairley, N. H. (1952). Cross-resistance between ‘‘Daraprim’’ and Proguanil. Br. Med. J. ii, 1255–1256. Rollo, I. M., Williamson, J., and Lourie, E. M. (1948). Acquired paludrine-resistance in Plasmodium gallinaceum. II. Failure to produce such resistance by prolonged treatment of latent infections. Ann. Trop. Med. Parasitol. 42, 241–248. Rosario, V. (1981). Cloning of naturally occurring mixed infections of malaria parasites. Science 212, 1037–1038. Ross, R. (1910). ‘‘The Prevention of Malaria.’’ E. P. Dutton & Company, New York. Russell, P. F., West, L. S., and Manwell, R. D. (1946). ‘‘Practical Malariology.’’ W. B. Saunders Company, Philadelphia. Salanti, A., Dahlback, M., Turner, L., Nielsen, M. A., Barfod, L., Magistrado, P., Jensen, A. T., Lavstsen, T., Ofori, M. F., Marsh, K., Hviid, L., and Theander, T. G. (2004). Evidence for the involvement of VAR2CSA in pregnancy-associated malaria. J. Exp. Med. 200, 1197–1203.

Strain Theory of Malaria: The First 50 Years

45

Sallenave-Sales, S., Daubersies, P., Mercereau-Puijalon, O., Rahimalala, L., Contamin, H., Druilhe, P., Daniel-Ribeiro, C. T., and Ferreira-da-Cruz, M. F. (2000). Plasmodium falciparum: A comparative analysis of the genetic diversity in malaria-mesoendemic areas of Brazil and Madagascar. Parasitol. Res. 86, 692–698. Sama, W., Owusu-Agyei, S., Felger, I., Dietz, K., and Smith, T. (2006). Age and seasonal variation in the transition rates and detectability of Plasmodium falciparum malaria. Parasitology 132, 13–21. Schofield, L., and Grau, G. E. (2005). Immunological processes in malaria pathogenesis. Nat. Rev. Immunol. 5, 722–735. Schwartz, R., Mansuy, M. M., and John, W. C. (1950). Current concepts on malaria. Arch. Intern. Med. 86, 837–856. Seaton, D. R., and Adams, A. R. D. (1949). Acquired resistance to proguanil in Plasmodium falciparum. Lancet ii, 323–324. Seaton, D. R., and Lourie, E. M. (1949). Acquired resistance to proguanil (paludrine) in Plasmodium vivax. Lancet i, 394–395. Sergent, E., Parrot, L., and Donatien, A. (1924). Une question de terminologie, immuniser et premunir. Bulletin de la Socie´te´ de Pathologie Exotique 17, 37–38. Sergent, E., Parrot, L., and Donatien, A. (1925). On the necessity of having a term to express the resistance of carriers of germs to superimposed infections. Trans. R. Soc. Trop. Med. Hyg. 18, 383–385. Shortt, H. E., and Garnham, P. C. C. (1948). The pre-erythrocytic development of Plasmodium cynomolgi and Plasmodium vivax. Trans. R. Soc. Trop. Med. Hyg. 41, 785–795. Shortt, H. E., Fairley, N. H., Covell, G., Shute, P. G., and Garnham, P. C. C. (1951). The pre-erythrocytic stage of Plasmodium falciparum. Trans. R. Soc. Trop. Med. Hyg. 44, 405–419. Shute, P. G. (1937). ‘‘Report on a third visit to Roumania for the Study of Malaria.’’ League of Nations Health Organisation Malaria Commission, Geneva (document 250). Shute, P. G. (1940). Failure to infect English specimens of Anopheles maculopennis var. atroparvus with certain strains of Plasmodium falciparum of tropical origin. J. Trop. Med. Hyg. 43, 175–178. Shute, P. G. (1946). Latency and long-term relapses in benign tertian malaria. Trans. R. Soc. Trop. Med. Hyg. 40, 189–200. Shute, P. G. (1951). Mosquito infection in artificially induced malaria. Br. Med. Bull. 8, 56–63. Shute, P. G. (1958). Thirty years of malaria-therapy. J. Trop. Med. Hyg. 61, 557–561. Shute, P. G., and Maryon, M. (1954). A contribution to the problem of strains of human plasmodium. Riv. Malariol. 33, 1–21. Sinden, R. E., Alavi, Y., and Raine, J. D. (2004). Mosquito-malaria interactions: A reappraisal of the concepts of susceptibility and refractoriness. Insect Biochem. Mol. Biol. 34, 625–629. Sinton, J. A. (1931). Studies in malaria, with special reference to treatment. Part XV. Does the strain of parasite influence cure? Indian J. Med. Res. 18, 845–853. Sinton, J. A. (1939). A summary of our present knowledge of the mechanism of immunity in malaria. J. Malaria Inst. India 2, 71–83. Sinton, J. A. (1940). Studies of infections with Plasmodium ovale. V. Trans. R. Soc. Trop. Med. Hyg. 33, 585–595. Smith, F. (1898). Malaria: Immunity: Absence of negro immunity: Variety. Br. Med. J. ii, 1807. Stephens, J. W. W. (1922). A new malaria parasite of man. Ann. Trop. Med. Parasitol. 16, 383–388. Stratman-Thomas, W. K. (1941). The infection in the intermediate host. Symptomatology: Vivax malaria. In ‘‘A Symposium on Human Malaria’’ (F. R. Moulton, ed.), pp. 183–189. American Association for the Advancement of Science, Washington DC. Taliaferro, W. H. (1949). Immunity to the malaria infections. In ‘‘Malariology’’ (M. F. Boyd, ed.), pp. 935–965. W. B. Saunders Company, Philadelphia.

46

F. Ellis McKenzie et al.

Thaithong, S., Beale, G. H., Fenton, B., McBride, J., Rosario, V., Walker, A., and Walliker, D. (1984). Clonal diversity in a single isolate of the malaria parasite Plasmodium falciparum. Trans. R. Soc. Trop. Med. Hyg. 78, 242–245. Thomson, J. G. (1931). The question of immunity in man to protozoal diseases. Proc. R. Soc. Med. 24, 499–513. Thomson, J. G. (1933). Immunity in malaria. Trans. R. Soc. Trop. Med. Hyg. 26, 483–514. Trager, W., and Jensen, J. B. (1976). Human malaria parasites in continuous culture. Science 193, 673–675. Trager, W., Tershakovec, M., Lyandvert, L., Stanley, H., Lanners, N., and Gubert, E. (1981). Clones of the malaria parasite Plasmodium falciparum obtained by microscopic selection: Their characterization with regard to knobs, chloroquine sensitivity, and formation of gametocytes. Proc. Natl. Acad. Sci. USA 78, 6527–6530. Viebig, N. K., Gamain, B., Scheidig, C., Lepolard, C., Przyborski, J., Lanzer, M., Gysin, J., and Scherf, A. (2005). A single member of the Plasmodium falciparum var multigene family determines cytoadhesion to the placental receptor chondroitin sulphate A. EMBO Rep. 6, 775–778. Vlachou, D., and Kafatos, F. C. (2005). The complex interplay between mosquito positive and negative regulators of Plasmodium development. Curr. Opin. Microbiol. 8, 415–421. Voller, A. (1964). Comments on the detection of malaria antibodies. Am. J. Trop. Med. Hyg. 13, 204–208. Wagner-Jauregg, J. (1922). The treatment of general paresis by inoculation of malaria. J. Nerv. Ment. Dis. 55, 369–375. Walliker, D. (1983). The genetic basis of diversity in malaria parasites. Adv. Parasitol. 22, 217–259. Watson, R. B. (1945). On the probability of soldiers with Pacific Plasmodium vivax malaria infecting Anopheles quadrimaculatus. J. Natl. Malar. Soc. 4, 183–188. Wellems, T. E., and Plowe, C. V. (2001). Chloroquine-resistant malaria. J. Infect. Dis. 184, 770–776. Wilson, D. B. (1936). Rural hyper-endemic malaria in Tanganyika Territory. Trans. R. Soc. Trop. Med. Hyg. 29, 583–618. World Health Organization (1963). ‘‘Terminology of Malaria and of Malaria Eradication.’’ Geneva. World Health Organization (1969). ‘‘Parasitology of Malaria.’’ Technical Report 433. Geneva. World Health Organization (1981). Malaria parasite strain characterization, cryopreservation, and banking of isolates: A WHO memorandum. Bull. World Health Organ. 59, 537–548. Yorke, W., and Macfie, J. W. S. (1924). Observations on malaria made during treatment of general paralysis. Trans. R. Soc. Trop. Med. Hyg. 18, 13–33. Young, M. D. (1944). Studies on the periodicity of induced Plasmodium vivax. J. Natl. Malar. Soc. 3, 237–240. Young, M. D., and Burgess, R. W. (1948). Studies on imported malarias 9. The comparative susceptibility of Anopheles quadrimaculatus and Anopheles maculopennis freeborni to foreign vivax malarias. J. Natl. Malar. Soc. 7, 134–137. Young, M. D., Ellis, J. M., and Stubbs, T. H. (1947). Some characteristics of foreign vivax malaria induced in neurosyphilis patients. Am. J. Trop. Med. 27, 585–596. Young, M. D., Eyles, D. E., and Burgess, R. W. (1949). Studies on imported malarias. 10. An evaluation of the foreign malarias introduced into the United States by returning troops. J. Natl. Malar. Soc. 7, 171–185. Young, M. D., Contacos, P. G., Stitcher, J. E., and Millar, J. W. (1963). Drug resistance in Plasmodium falciparum from Thailand. Am. J. Trop. Med. Hyg. 12, 305–314. Yount, E. H., Jr., and Coggeshall, L. T. (1949). Status of immunity following cure of recurrent vivax malaria. Am. J. Trop. Med. Hyg. 29, 701–705. Zuckerman, A. (1964). The antigenic analysis of plasmodia. Am. J. Trop. Med. Hyg. 13, 209–213.

CHAPTER

2 Advances and Trends in the Molecular Systematics of Anisakid Nematodes, with Implications for their Evolutionary Ecology and Host–Parasite Co-evolutionary Processes Simonetta Mattiucci* and Giuseppe Nascetti†

Contents

1. Introduction 2. Molecular Systematics of Anisakid Nematodes 2.1. Current methods for anisakid identification 3. The Current Taxonomy 3.1. The current taxonomy of Anisakis 3.2. The current taxonomy of Pseudoterranova decipiens (sensu lato) 3.3. The current taxonomy of Contracaecum species from pinnipeds 4. Phylogenetic Systematics of Anisakid Nematodes 4.1. Genetic relationships between Anisakis spp. 4.2. Genetic relationships between Pseudoterranova spp. 4.3. Genetic relationships between Contracaecum spp.

49 51 54 55 55 76 83 91 92 95 96

* Department of Public Health Sciences, Section of Parasitology, ‘‘Sapienza’’—University of Rome, {

P.le Aldo Moro, 5, 00185 Rome, Italy Department of Ecology and Sustainable Economic Development—Tuscia University—Via S. Giovanni Decollato, 1, 01100 Viterbo, Italy

Advances in Parasitology, Volume 66 ISSN 0065-308X, DOI: 10.1016/S0065-308X(08)00202-9

2008 Elsevier Ltd. All rights reserved.

#

47

48

Simonetta Mattiucci and Giuseppe Nascetti

5. Genetic Differentiation in Anisakids 5.1. Genetic differentiation at interspecific level 5.2. Genetic differentiation at the intraspecific level and gene flow 6. Host–Parasite Cophylogeny 7. Host Preference, Ecological Niche and Competition 8. Anisakids as Biological Indicators 8.1. Anisakis spp. larvae as biological tags of fish stocks 8.2. Anisakids as indicators of trophic web stability and habitat disturbance of marine ecosystems 9. Conclusions and Identification of Gaps in Our Knowledge of Anisakids to be Filled by Future Research 9.1. Molecular systematics 9.2. Identification of human infections 9.3. Molecular ecology and life cycle 9.4. Host–parasite co-evolutionary aspects 9.5. Genetic variability of anisakids as an indicator of habitat disturbance Acknowledgements References

Abstract

100 100 106 108 114 119 120 123

130 131 132 132 133 134 137 137

The application of molecular systematics to the anisakid nematodes of the genera Anisakis, Pseudoterranova and Contracaecum, parasites of aquatic organisms, over the last two decades, has advanced the understanding of their systematics, taxonomy, ecology and phylogeny substantially. Here the results of this effort on this group of species from the early genetic works to the current status of their revised taxonomy, ecology and evolutionary aspects are reviewed for each of three parasitic groups. It has been shown that many anisakid morphospecies of Anisakis, Contracaecum and Pseudoterranova include a certain number of sibling species. Molecular genetic markers provided a rapid, precise means to screen and identify several species that serve as definitive and intermediate and or/paratenic hosts of the so far genetically characterized species. Patterns of differential distribution of anisakid nematodes in various definitive and intermediate hosts are presented. Differences in the life history of related species can be due both to differential host–parasite co-adaptation and co-evolution, and/or to interspecific competition, that can reduce the range of potential hosts in sympatric conditions. Phylogenetic hypotheses attempted for anisakid nematodes and the possible evolutionary scenarios that have been proposed inferred from molecular data, also with respect to the phylogeny of their hosts are presented for the parasite–host associations Anisakis-cetaceans and

Anisakid Nematodes and Host–Parasite Co-evolutionary Processes

49

Contracaecum-pinnipeds, showing that codivergence and hostswitching events could have accompanied the evolution of these groups of parasites. Finally, examples in which anisakid nematodes recognized genetically at the species level in definitive and intermediate/paratenic hosts from various geographical areas of the Boreal and Austral regions and their infection levels have been used as biological indicators of fish stocks and food-web integrity in areas at high versus low levels of habitat disturbance (pollution, overfishing, by-catch) are presented.

1. INTRODUCTION Adult nematodes of the genera of most species of the anisakid nematodes Anisakis, Dujardin, 1845, Pseudoterranova Krabbe, 1878 and Contracaecum Railliet and Henry, 1913 are parasites of the alimentary tract of aquatic vertebrates. They display indirect life cycles in aquatic ecosystems and involve various hosts at different levels in food webs. Marine mammals (cetaceans and pinnipeds) and fish-eating birds serve as definitive hosts; fish, squids and other invertebrates serve as intermediate or paratenic hosts; and crustaceans serve as first intermediate hosts. In humans, several larval anisakid nematodes cause the zoonotic disease, presently known as ‘anisakidosis’ or ‘anisakiosis’, when consumed in raw or undercooked fish. Anisakis is considered as the most important anisakid genus with regard to human infection (Audicana et al., 2002, and references therein), but some species of Pseudoterranova also have been implicated in human infections (Adams et al., 1997; Oshima, 1987), and a few species of Contracaecum are potentially infective (Vidal-Martinez et al., 1994). This zoonosis has a history starting from the first report (Van Thiel, 1960) of a larval nematode from herring in the gastro-intestinal tract of humans in the Netherlands. Some aspects of human anisakidosis have recently been reviewed by Audicana et al. (2002). Because of the existence of several excellent reviews by others authors on the history and clinical aspect of the anisakidosis (Audicana et al., 2002; Chai et al., 2005; Umehara et al., 2007), this introduction will not summarize the major events that mark the history of anisakid nematodes as causative agents of anisakidosis, but instead will highlight major landmarks involved in our understanding of their taxonomic status and ecology based on genetic markers. This field has had a major impact during the last 20 years on our knowledge of these parasites, including their host-specificity, geographical range and the possible identification of human cases of anisakidosis. Starting from the knowledge summarized in the exhaustive revision by Smith and Wootten (1978), this review treats the explosion in the

50

Simonetta Mattiucci and Giuseppe Nascetti

literature that has accompanied research activities on these nematodes and resulted in a series of discoveries over the last 25 years on the systematics and ecology of anisakid nematodes using molecular genetic tools applied to their taxonomy. Indeed, based on the advances in our understanding of the basic biology of Anisakis, reviewed by Smith and Wootten (1978), this introduction starts from their concluding remarks highlighting knowledge gaps and perspectives for investigations over the next 30 years. One major gap concerned the identification of species. Indeed, in the case of the species belonging to the genus Anisakis, they wrote: ‘. . . despite Davey’s revision (1971), there are still taxonomic problems, including the distinction between A. simplex and A. typica . . .. . . The generic diagnosis of Anisakis should be re-examined . . .. Anisakis larva (I) from North Atlantic waters has been cultured in vitro and shown to develop into A. simplex: There is need to culture the other ‘‘larval types’’ in order to confirm that they do, in fact, represent Anisakis and, if so, to determine which species they represent . . ..’. In 1978, Smith and Wootten underlined the major gap in our knowledge of nematode biology at that date, as represented by the genus Anisakis but also relevant to other anisakid species, that is, the identification of biological species. Indeed, species identification, based on morphological characters only, is difficult for adults, but even more difficult for larval stages. The need for species identification was especially important for larval stages of Anisakis because they have been implicated as causative agents of human anisakidosis. Indeed, Smith and Wootten (1978) concluded their review as follows: ‘. . . in these circumstances there is need for reliable methods of differential diagnosis. There is much scope for future investigation’. Thus, this ‘future investigation’ started with the application of molecular genetic methodologies in an attempt to address the systematics of these anisakids. The prospect of assessing anisakid nematode biodiversity based on molecular genetic markers as the preferred diagnostic tools seemed promising because the unambiguous identification of specimens causing human disease (anisakidosis) is essential for a proper epidemiological survey. The power of resolution of molecular genetic methodologies in detecting anisakid species revolutionized the taxonomy of these species during the following 25 years. Most descriptions of parasite species conformed with what can be regarded as the ‘morphological or typological, species concept’. Because genetic speciation is not always accompanied by corresponding morphological change, the actual number of biological species is likely to be greater than the current tally of nominal species, most of which are delineated based on morphological grounds. Detecting biological species of anisakid nematodes challenged parasitologists via an expected genetic variation and heterogeneity within the nominal species and has led to the definition of anisakid species according to the Mayr’s (Mayr, 1963) ‘biological species concept’ (BSC).

Anisakid Nematodes and Host–Parasite Co-evolutionary Processes

51

The aim of this review is to gather together the many and varied aspects concerning the biology and evolutionary ecology of those anisakid species detected genetically and recognized as members of Anisakis, Pseudoterranova and Contracaecum, including (a) species presently accepted as senior synonyms based on the application of different molecular genetic markers; (b) current molecular genetic approaches to identify anisakid species; (c) ecological evidence relating to the geographical distribution of detected species, their host preferences and life cycles; (d) the use of anisakid nematodes as biological tags for fish stock identification in a multidisciplinary approach; (e) correlation between the values of genetic variability and the levels of parasitic infection in definitive and intermediate/paratenic hosts as indicators of the integrity of marine food webs; (f) estimates of genetic divergence at intraspecific and interspecific levels; and (g) estimates of their genetic relationships based on different clustering approaches inferred from different molecular genetic data sets and their use to infer phylogenetic hypothesis and possible co-evolutionary events with their definitive hosts.

2. MOLECULAR SYSTEMATICS OF ANISAKID NEMATODES The nematode superfamily Ascaridoidea contains
52 genera; species are mainly parasites of the alimentary tract of vertebrates. Several are of medical and veterinary concern, and some are of economic significance. The classification scheme for members of the Anisakidae proposed by Hartwich (1974) is based largely on features of the ‘excretory system’ and alimentary tract. According to the ‘systematic keys’ of Hartwich (1974), the family Anisakidae contains the subfamilies Anisakinae and Contracaecinae. The Anisakinae includes (cf. Hartwich, 1974) the genera Anisakis and Pseudoterranova plus several others. The Contracaecinae includes three genera, Contracaecum, Phocascaris and Galeiceps. Species of Anisakis, Pseudoterranova, Contracaecum and Phocascaris have aquatic life cycles and homeothermic final hosts. Anisakid nematodes have been extensively studied with respect to their alpha-taxonomy (Berland, 1961, 1964; Bruce and Cannon, 1990; Davey, 1971; Deardoff and Overstreet, 1979; Fagerholm, 1989, 1991; Fagerholm and Gibson, 1987; Gibson, 1983; Osche, 1963; Sprent, 1977, 1978, 1979, 1983) and life cycles (Huizinga, 1967; Klo¨ser et al., 1992; Kie and Fagerholm, 1995). Much of the controversy in the systematics of anisakid species, when dealing with identifications based solely on morphological differences, revolved around their confused taxonomic position. Indeed, morphological characters of taxonomic significance in this group are very few (i.e. features of the excretory system, alimentary canal, number and

52

Simonetta Mattiucci and Giuseppe Nascetti

distribution of male caudal papillae, position of the vulva and length of the spicules) and are applicable to adult specimens only. Furthermore, these are often only relevant to male individuals, making identification difficult for many worms at the species level. In recent decades, molecular genetic techniques have been used to demonstrate that structural features are not always adequate for recognizing ‘true’ species within Anisakis, Pseudoterranova and Contracaecum. In the last two decades, the diversity of anisakid species has increased due to the detection by genetic markers of several sibling species with reproductively isolated gene pools that are morphological very similar, and thus correspond to the ‘biological species’. There are now morphospecies, or species complexes, based on previously recognized cosmopolitan species (sensu lato), that may comprise several recognized species. This solved one of the major problems in the systematics of anisakid nematodes: the occurrence of the parallelism and convergence of morphological features, which confound the systematic value of some morphological criteria and often accompany a high genetic and ecological divergence between the species. The lack of morphological differences in these parasites may be due to factors such as similar selection pressures causing the conservation of morphology; consequently, some morphological characters have little or no taxonomic value because of the evolutionary co-adaptation of these endoparasites to the stable habitat represented by their definitive hosts. Indeed morphospecies may appear to have multiple host species, that is, parasite populations isolated in their hosts have diverged genetically but have conserved morphological features. Moreover, species identification based on morphological characters only makes identification very difficult, especially for larval stages that lack reliable diagnostic features at the species level. Thus, the inconsistency in morphological characteristics of anisakid nematodes impeded the development of a credible scheme for their taxonomy. This prompted the need to classify these nematodes using genetic and/or biochemical methods. The assessment of anisakid nematodes biodiversity based on molecular genetic markers as preferable tools for specific diagnosis was an important prospect because the unambiguous identification of those anisakids with a zoonotic potential is an essential requirement for a proper epidemiological survey. Initial attempts to apply population genetics to the study of genetic variation among large samples of anisakids collected from different intermediate/paratenic and definitive hosts in nature employed the use of multilocus allozyme electrophoresis (MAE) (19–24 enzyme loci). These tools revealed the existence of high genetic heterogeneity within certain anisakid morphospecies, such as those of Anisakis, Pseudoterranova and Contracaecum (Bullini et al., 1986; Mattiucci et al., 1986, Nascetti et al., 1986;

Anisakid Nematodes and Host–Parasite Co-evolutionary Processes

53

Paggi et al., 1991). The species concept (BSC) (Mayr, 1963) was well supported by the application of allozyme markers for several anisakid species. Indeed, after the first such application, the diversity of species belonging to Anisakis, Pseudoterranova and Contracaecum quickly increased after the detection of several sibling species and leads to the discovery and description of several new species. Reproductive isolation and absence of gene flow were demonstrated by these allozymes between sympatric and allopatric sibling species, establishing their specific status (Mattiucci et al., 1997, 2001, 2003, 2005; Nascetti et al., 1993; Paggi et al., 1991). Allozyme markers have allowed us to (a) genetically characterize different species of anisakid nematodes, (b) estimate their genetic differentiation, (c) establish their genetic relationships, and (d) identify their larval stages which lack morphological characters (Arduino et al., 1995; Bullini et al., 1986, 1994, 1997; Mattiucci and Nascetti, 2006; Mattiucci et al., 1986, 1997, 1998, 2001, 2002a, 2003, 2004, 2005, 2006, 2007a, 2008a,b,c; Nascetti et al., 1986, 1993; Orecchia et al., 1986a, b, 1994; Paggi and Bullini, 1994; Paggi et al., 1991, 1998b, 2000, 2001). The introduction of the polymerase chain reaction (PCR)-derived molecular methodologies later confirmed taxonomic decisions involving anisakid species previously based on allozyme markers. Reference individuals initially characterized by allozymes have been used to develop DNA-based approaches for species identification, such as PCR-RFLP and direct sequencing of ITS rDNA or mitochondrial DNA. Ascaridoid classifications and inferred patterns of character evolution have been investigated previously, and some evolutionary hypotheses for representative ascaridoids, some of which are anisakids, have been proposed based on phylogenetic methods, including ribosomal DNA sequence data and comparative analysis of morphological and life-history characters (Nadler, 1992, 1995, 2005; Nadler and Hudspeth, 1998; Zhu et al., 2000a). Results from these studies identified some consistent, putative, shared-derived morphological features, strongly suggesting that some morphological features represent ancestral states or highly homoplastic characteristics (Nadler and Hudspeth, 1998). Phylogenetic analysis indeed provided a new perspective for the delimitation of anisakid sibling species, including hierarchical relatedness and relative rates of evolution. An evolutionary perspective provides a conceptual approach to view species as independent evolutionary lineages and provides another approach for delimitating species (Adams, 1998; Nadler, 2002, 2005). Indeed, based on phylogenetic DNA analysis, sibling anisakid species have been confirmed by methods that can test the hypothesis of lineage independence analysing many individual specimens and sometimes detecting new genotypes and species (D’Amelio et al., 2000, 2007; Mattiucci et al., 2008b, c; Nadler et al., 2000; 2005; Valentini et al., 2006).

54

Simonetta Mattiucci and Giuseppe Nascetti

2.1. Current methods for anisakid identification Except for some morphological differences between anisakid species (Mattiucci et al., 1998, 2005, 2008c; Paggi et al., 1998b, 2000), all the species genetically characterized are to date indistinguishable at all developmental stages (larvae and adults); consequently, only molecular genetic markers can be used reliably to identify these species. Of all the recently developed tools, and despite the useful PCR-derived molecular markers, allozymes continue to remain the best methodological approach for demonstrating reproductive isolation between anisakids. Starting from the study of genetic variation at several enzyme loci, the main argument in favour of using allozymes is the ability to test (using the Hardy– Weinberg equilibrium) for reproductive isolation of anisakid populations among large numbers of individuals. Accordingly, allozymes provide the best choice for determining molecular genetic markers applied to the systematics of anisakids and for confirming the BSC. Based on allozymes diagnostic for different anisakid taxa, easy and rapid identification of large numbers of individuals can be performed; this method is particularly valuable for identifying larval individuals collected from several intermediate/paratenic hosts and often in mixed infections. Accordingly, such identifications have been demonstrated to be very informative tools for answering epidemiological questions involving geographical range, host preference and life cycles. Moreover, because numerous allozymes (20–24 enzyme loci) have been applied to thousands of anisakid individuals, they have contributed greatly to our knowledge of the genetic diversity of anisakid populations collected from various ecosystems in the Boreal and Austral hemispheres (see also Section 8.2). However, the tool is limited to frozen-preserved or fresh individuals. This constraint has been resolved by DNA-based diagnostic techniques, which have the advantage of also being able to use alcohol or formalin-preserved specimens. In contrast with allozymes, the DNAbased techniques have increased our ability to study phylogenetic relationships between related anisakids based on the evolutionary lineage concept (Adams, 1998). The PCR-DNA molecular derived methodologies so far applied to the systematics of anisakid nematodes include PCRrestriction fragment length polymorphism (PCR-RFLPs of ITS-DNA) (D’Amelio et al., 2000; Kijewska et al., 2002; Pontes et al., 2005), singlestrand conformational polymorphism SSCP-DNA (ITS) of PCR products (Zhu et al., 1998, 2000b, 2007), direct sequencing of PCR-amplified DNA of 28S (LSU) and complete internal transcribed spacer (ITS-1, 5.8S, ITS-2) ribosomal DNA (Hu et al., 2001; Li et al., 2005; Nadler et al., 2000, 2005; Zhu et al., 1998, 2000b, 2001, 2002), mitochondrial cytochromoxidase b (mtDNA cytb) (Mattiucci et al., 2003) and mitochondrial cytochromoxidase 2 (mtDNA cox2) sequence analyses (Mattiucci and Nascetti, 2006;

Anisakid Nematodes and Host–Parasite Co-evolutionary Processes

55

Mattiucci et al., 2008b, c; Valentini et al., 2006). PCR-DNA markers have so far confirmed the existence of the sibling species previously detected by allozymes, establishing their taxonomic status. The only exception is represented by the two Antarctic sibling species of the Contracaecum osculatum complex (i.e. C. osculatum D and C. osculatum E), where PCR-DNA markers based on ITS-rDNA (Zhu et al., 2000b) and three mitochondrial DNA genes (i.e. mtDNA cox1, ssrRNA and lsrDNA) (Hu et al., 2001) were unable to identify species previously detected by allozyme markers. In contrast, the existence of the two sibling species, C. osculatum D and C. osculatum E, first detected by allozymes, was confirmed by sequencing the mtDNA cox2 gene (Mattiucci et al., 2008b). A scheme of allozymes and PCR-derived methods which can be used routinely to identify a single individual belonging to genetically characterized species and gene pools of anisakids as belonging to the genera Anisakis, Pseudoterranova and Contracaecum is provided in Tables 2.1–2.3, respectively.

3. THE CURRENT TAXONOMY This section summarizes the current taxonomy of anisakid species of the genera Anisakis, Pseudoterranova and Contracaecum (here considering only those species maturing in pinnipeds) which have been genetically characterized to date. They can be recognized, at any life-history stage, by different molecular genetic approaches, as reported in Tables 2.1–2.3. A synopsis of each recognized anisakid species, including data on both the definitive and intermediate hosts and the geographical range, is also presented.

3.1. The current taxonomy of Anisakis The taxonomy of Anisakis species has traditionally relied on adult morphology. According to Davey (1971), the primary characters are the length and shape of the ventriculus, length and shape of the male spicules, and arrangement of the male caudal papillae. According to Berland (1961), larval morphological features (i.e. length of ventriculus and presence/ absence of caudal spine) could distinguish Anisakis type I and Anisakis type II larvae. In the revision by Davey (1971), the generic diagnosis for Anisakis states: ‘three lips each bearing a bilobed anterior projection which carries the single dentigerous ridge; interlabia absent; excretory gland with duct opening between ventrolateral lips; oesophagus with anterior muscular portion (proventriculus) and posterior ventriculus, the latter being oblong and sometimes sigmoid or else as broad as long; no oesophageal appendix or intestinal caecum; vulva in middle or first third of the body, spicules of male unequal; preanal papillae numerous; postanal papillae including a group of three or four

56 TABLE 2.1 Molecular genetic markers for the identification of the species of Anisakis Molecular genetic marker

Method

Identified speciesa

References

22 allozyme loci 20 allozyme loci 25 allozyme loci 20 allozyme loci 5.8S rDNA, ITS rDNA 18S and 5.8S rDNA, ITS rDNA 22 allozyme loci

MAE MAE MAE MAE PCR-RFLP; SSCP PCR-RFLP

Nascetti et al., 1986 Mattiucci et al., 1997 Paggi et al., 1998b Mattiucci et al., 1998 Zhu et al., 1998 D’Amelio et al., 2000

21 allozyme loci

MAE

20 allozyme loci

MAE

18S, 28S and 5.8S rDNA, ITS rDNA ITS rDNA

PCR and sequencing

Asstr, Ape Asstr, AsC, Ape Azi, Ape, AsC, Aph Asstr Asstr Ape, AsC, Aty, Azi, Aph, Abr, Asc Abr, Aph, Azi, Ape, AsC, Aph Aty, Asstr, Ape, Azi, Ape, AsC, Aph Apa, Asstr, Ape, Aty, Az, Aph, Abr AsC, Ape, Asp Asstr, Ape, Aty, Azi, Aph, AspA

Pontes et al., 2005

MAE

PCR-RFLP

Mattiucci et al., 2001 Mattiucci et al., 2002a Mattiucci et al., 2005 Nadler et al., 2005

a

20 allozyme loci

MAE

Mattiucci and Nascetti, 2006 Valentini et al., 2006

PCR-RFLP

Asp, Asstr, Ape, AsC, Aty, Azi, Aph, Ab, Apa Asstr, Ape, AsC, Aty, Azi, Aph, Abr, Apa, Asp Asstr, Ape

Mitochondrial cytochrome c-oxidase subunit 2 (mtDNA cox2) ITS rDNA, 5.8S rDNA, mitochondrial cytochrome c-oxidase subunit 1 (mtDNA cox1) ITS rDNA, 5.8S rDNA Mitochondrial cytochrome c-oxidase subunit 1 (mtDNA cox1), NADH dehydrogenase subunit 1 ITS-2 rDNA ITS rDNA

PCR and sequencing

PCR-RFLP PCR and sequencing

Asstr, Ape Asstr

Abe et al., 2006 Cross et al., 2007

PCR-SSCP Multiplex-PCR

Ape, Aty Asstr, Ape, Aph

Zhu et al., 2007 Umehara et al., 2008

Umehara et al., 2006

Codes for Anisakis spp.: Asstr, A. simplex (s.s.); Ape, A. pegreffii; AsC, A. simplex C; Aty, A. typica; Azi, A. ziphidarum; Aph, A. physeteris; Abr, A. brevispiculata; Apa, A. paggiae; Asp, Anisakis sp.; Asc, A. schupakovi; AspA, Anisakis sp. A.

57

58

Simonetta Mattiucci and Giuseppe Nascetti

TABLE 2.2 Molecular genetic markers for the identification of the Pseudoterranova decipiens species complex Molecular genetic marker

Identified speciesa

References

MAE

Pdss, Pkr, Pbu

Paggi et al., 1991

MAE

Pca

16 allozyme loci

MAE

20 allozyme loci ITS rDNA

MAE

Pdss, Pkr, Paz, Pbu, PdE Pbu, Paz, Pdss, Pkr Pdss, Pkr, Paz, Pbu, Pca, PdCa1 Pdss, Pkr, Paz, Pbu, Pca, PdCa1

George-Nascimento and Llanos, 1995; George-Nascimento and Urrutia, 2000 Bullini et al., 1997

16 allozyme loci 9 allozyme loci

18S, 28S and 5.8S rDNA, ITS rDNA a

Method

PCR-SSCP

PCR and sequencing

Mattiucci et al., 1998 Zhu et al., 2002

Nadler et al., 2005

Codes for Pseudoterranova spp.: Pdss, P. decipiens (s.s.) (=P. decipiens B); Pkr, P. krabbei (=P. decipiens A); Pbu, P. bulbosa (=P. decipiens C); Paz, P. azarasi (=P. decipiens D); PdE, P. decipiens E; Pca, P. cattani; PdCa1, P. decipiens from Chaenocephalus aceratus.

pairs set close to the tip of the tail on ventral side’. After a critical revision of the 21 included species, Davey (1971) concluded that many of the species had to be considered as junior synonyms of three main species, the only ones accepted in his revision, that is, A. simplex (Rudolphi, 1809, det. Krabbe, 1878) with 10 synonyms, A typica (Diesing, 1860) with one synonym and A. physeteris (Baylis, 1923) with three synonyms. He also retained four others as species inquirendae because of the lack of sufficient data: A. dussurmierii (van Beneden, 1870) reported from a dolphin in the Indian Ocean, A. schupakovi Mozgovoi, 1951 from the Caspian seal, Pusa caspica (later redescribed and accepted by Delyamure et al., 1964), A. alexandri Hsu¨ and Hoeppli, 1933 from Sotalia sinensis and A. insignis (Diesing, 1851) from Inia geoffrensis in South America rivers (redescribed and accepted by Petter, 1972). According to Davey’s revision, the only morphological characters of systematic value in recognizing the taxonomic status of the species of Anisakis were the length and shape of ventriculus, length and shape of the

TABLE 2.3

Molecular genetic markers for the identification of the species of Contracaecum and Phocascaris from pinnipeds

Molecular genetic marker

Method

Identified speciesa

References

17 allozyme loci 24 allozyme loci

MAE MAE

Nascetti et al., 1993 Orecchia et al., 1994

25 allozyme loci

MAE

20 allozyme loci 28S rDNA

MAE PCR and sequencing

ITS rDNA ITS rDNA 18S and 28S rRNA, mitochondrial cytochrome c-oxidase subunit 1 (mtDNA cox1) 18 allozyme loci

PCR-SSCP PCR-RFLP; PCR-SSCP PCR-SSCP

CoA, CoB, Coss CoD, CoE, CoA, CoB, Coss Cra, CoA, CoB, Coss, CoD, CoE CoA, CoB CoA, CoB, Coss, Cba, Crad, Cmir, Pcy, Pph CoA, CoB, Coss, Cbai Cogm CoA, CoB, Coss, Cbai

MAE

Cogm, Cmar

Arduino et al., 1995 Mattiucci et al., 1998 Nadler et al., 2000 Zhu et al., 2000 Zhu et al., 2001 Hu et al., 2001

Mattiucci et al., 2003 (continued)

TABLE 2.3 (continued)

a

Molecular genetic marker

Method

Identified speciesa

References

18S, 28S and 5.8S rDNA, ITS rDNA

PCR and sequencing

Nadler et al., 2005

20 allozyme loci

MAE

Mitochondrial cytochrome c-oxidase subunit 2 (mtDNA cox2)

PCR and sequencing

CoA, CoB, Coss, Crad, Cmir, Cbai, Cmar, Cogm, Pcys, Pph CoA, CoB, Coss, CoD, CoE, Cbai, Crad, Cmir, Cmar, Cogm, Pcys CoA, CoB, Coss, CoD, CoE, Cbai, Crad, Cmir, Cmar, Cogm, Pcys

Mattiucci et al., 2008b

Mattiucci et al., 2008b

Codes for Contracaecum spp.: CoA, C. osculatum A; CoB, C. osculatum B; Coss, C. osculatum (s.s.) (=C. osculatum C); CoD, C. osculatum D; CoE, C. osculatum E; Cbai, C. baicalensis; Cmar, C. margolisi; Crad, C. radiatum; Cogm, C. ogmorhini (s.s.); Cmir, C. mirounga; Pph, Phocascaris phocae; Pcys: Ph. Cystophorae.

Anisakid Nematodes and Host–Parasite Co-evolutionary Processes

61

male spicules, arrangement of the male caudal papillae and position of the vulva in females. Morphological features such as the length of the ventriculus and presence/absence of caudal spine were proposed for the recognition of larval stages of Anisakis, but these are too inconsistent and inaccurate for use at the specific level (Berland, 1961; Shiraki, 1974). Despite Davey’s valuable discussion, when considering the main morphological characters of the species Anisakis, instability of their systematics, including their identification at the species level (especially in the case of females and larvae), remained. The resolution of some of the taxonomic issues, highlighted in this critical revision, has been achieved during the last 20 years facilitated by the adoption of new methodologies, allozymes first and DNA-based molecular methodologies later. Starting from that critical revision, and based on our growing understanding on the systematic of species Anisakis, we can state that Davey’s revision failed in considering only three species as valid and several species as synonyms, as some of the latter have been subsequently validated by genetic methodology (Mattiucci et al., 2001). However, this revision anticipated, in some way, a future cladistic analysis by indicating morphological characters likely to be useful for the systematics of the species belonging to Anisakis following their genetic identification. In addition, in his revision, Davey did not consider Skryabinisakis Mosgovoi, 1951, proposed for the species A. physeteris, as an accepted genus. However, the high level of genetic differentiation of both A. physeteris, A. brevispiculata and A. paggiae (see Mattiucci et al., 2005) from those species of Anisakis studied genetically using different nuclear and mitochondrial markers (Mattiucci and Nascetti, 2006; Mattiucci et al., 2001; Valentini et al., 2006) indicates that Anisakis is indeed polyphyletic and highly heterogeneous. Today, the existence of two main clades within Anisakis is clearly inferred from the phylogenetic analysis based on nuclear data sets from allozymes and mtDNA cox2 sequence analysis of all the genetically characterized species (Mattiucci and Nascetti, 2006; Mattiucci et al., 2005; Valentini et al., 2006). One clade encompasses species with the larval stage known as Anisakis type I (sensu Berland, 1961) and the second sharing the larval morphology of Anisakis type II (sensu Berland, 1961) (see Mattiucci et al., 2005, 2007a; Orecchia et al., 1986a). The first clade includes the species of the A. simplex complex [i.e. A. simplex (sensu stricto), A. pegreffii and A. simplex C], A. typica, A. ziphidarum and Anisakis sp. The second includes the species A. physeteris, A. brevispiculata and A. paggiae (see Mattiucci et al., 2005; Valentini et al., 2006). The clade including A. simplex (s.s.) and A. simplex C and A. pegreffii is also well supported by a phylogenetic analysis inferred from ITS rDNA sequence data sets (Nadler et al., 2005). This analysis is congruent with both allozymes and mtDNA cox2 analysis in depicting A. physeteris + A. brevispiculata as the sister group to the remaining Anisakis spp.

62

Simonetta Mattiucci and Giuseppe Nascetti

3.1.1. Anisakis spp. included in clade I According to the genetic data, six species can currently be included in this clade: three species of the A. simplex complex [i.e. A. simplex (s.s.), A. pegreffii and A. simplex C], A. typica, A. ziphidarum and a new gene pool referred to as Anisakis sp. (see Valentini et al., 2006, unpublished data). As far as is known, all of these species have the larval form known as type I (sensu Berland, 1961). A synopsis of the ecological aspects of each species, including their known geographical distribution and both definitive and intermediate/paratenic hosts, is presented below. A. simplex (Rudolphi, 1809) (sensu stricto) (see Nascetti et al., 1986): A. simplex (s.s.) is widespread between 35 N and the Arctic Circle; it is present in both the western and eastern Atlantic and Pacific Oceans (Abe et al., 2005, 2006; Abollo et al., 2001; Mattiucci et al., 1997, 1998; Nadler et al., 2005; Paggi et al., 1998a; Umehara et al., 2006, 2008) (Fig. 2.1). The southern limit of this species in the north east Atlantic Ocean is the waters around the Gibraltar area. A. simplex (s.s.) is also occasionally present in western Mediterranean waters due to the migration of pelagic fish species into the Alboran Sea from the Atlantic (Mattiucci and Nascetti, 2006; Mattiucci et al., 2004, 2007a). A. simplex (s.s.) has been so far genetically recognized from nine species of cetacean hosts. Several squid and fish species have been found harbouring larvae of this species throughout its geographical range (Table 2.5). A sympatric area between A. simplex (s.s.) and A. pegreffii has been identified along the Spanish and Portuguese Atlantic coasts (Abollo et al., 2001; Marques et al., 2006; Mattiucci et al., 1997, 2004, 2007a; Pontes et al., 2005), in the Alboran Sea (Mattiucci et al., 2004, 2007a) and recently also in Japanese Sea waters (Umehara et al., 2006). A. simplex (s.s.) also occurs with A. simplex C in the eastern Pacific Ocean, where it has been identified in definitive and intermediate/paratenic hosts (Mattiucci et al., 1997, 1998, unpublished data; Paggi et al., 1998c) (Tables 2.4 and 2.5; Fig. 2.1). Although it has sympatric and syntopic occurrences in mixed infections at both larval and adult stage with other Anisakis species (Mattiucci et al., 2004, 2005, 2007a), reproductive isolation between A. simplex (s.s.) and both A. pegreffii and A. simplex C was proved by the lack of adult F1 hybrids and/or backcross genotypes clearly demonstrated at the nuclear level by allozyme markers (Mattiucci et al., 1997, 2005). Only a few F1 hybrid larval individuals A. pegreffii–A. simplex (s.s.), of the thousands identified, were detected by allozymes in some fish host from the sympatric area off the Atlantic Iberian coast (Mattiucci et al., 2004). However, back-crossing of F1 hybrids with parental species has not been detected. Based on PCR-RFLP of the ITS, Umehara et al. (2007) recognized A. simplex (s.s.) as the main source of infections in humans in Japan.

60⬚

i olar C Ar tic P

rcle

30⬚

0⬚

30⬚

Antarc

60⬚

tic Po

60⬚

120⬚ A. pegreffii

A. simplex (s.s. )

C. osculatum (s.s.) P. decipiens (s.s.) Ph. phocae

A. simplex C

C. osculatum A P. krabbei

A. typica

C. osculatum B

P. bulbosa

P. azarasi

A. physeteris

C. o. baicalensis P. decipiens E

0⬚ A. brevispiculata C. osculatum D

lar Cir

cle

60⬚ A. ziphidarum C. osculatum E

180⬚

120⬚ A. paggiae C. radiatum

Anisakis sp. C. mirounga

C. ogmorhini (s.s)

C. margolisi

P. cattani

Ph. cystophorae

FIGURE 2.1 World map showing the so far known distribution areas of anisakid species of Anisakis (□), Pseudoterranova (△), Contracaecum (○) and Phocascaris (?). The geographical areas indicated are related to the sampling localities for their definitive and intermediate hosts.

64

TABLE 2.4

Definitive hosts so far detected, by molecular genetic markers, for the Anisakis spp.

Cetaceans Balaenopteridae Balaenoptera acutorostrata Delphinidae Delphinus delphis Globicephala melaena Globicephala macrorhynchus Lagenorhynchus albirostris Lissodelphis borealis Orcinus orca Pseudorca crassidens Stenella coeruleoalba Tursiops truncatus Sotalia fluviatilis Stenella attenuata Stenella longirostris Steno bredanensis

A. simplex (s.s.)

A. pegreffii A. simplex C A. typica

A. ziphidarum

Anisakis sp. A. physeteris

A. brevispiculata A. paggiae

NEA

–

–

–

–

–

–

–

–

IC IC, SA

IC, WM –

– SA

– –

– –

– –

– –

– –

– –

–

–

–

FL

–

–

–

–

–

NEA

–

–

–

–

–

–

–

–

–

–

NEP

–

–

–

–

–

NEP NEP

–

– NEP

–

– –

– –

– –

– –

– –

IC

WM

–

EM

–

–

–

–

–

–

CM, SA

–

FL, CS

–

–

–

–

–

– – –

– – –

– – –

BR FL, CS BR

– – –

– – –

– – –

– – –

– – –

–

–

–

CS

–

–

–

–

–

Kogiidae Kogia breviceps Kogia sima Monodontidae Delphinapterus leucas Neobalaenidae Caperea marginata Phocoenidae Phocoena phocoena Physeteridae Physeter catodon Ziphiidae Mesoplodon densirostris Mesoplodon europaeus Mesoplodon grayi Mesoplodon mirus Ziphius cavirostris

– –

– –

– –

– –

– –

– –

– –

SA, IC, FL –

SA, FL FL

NWA

–

–

–

–

–

–

–

–

–

SA

–

–

–

–

–

–

–

NEP

–

–

–

–

–

–

–

–

–

–

–

–

–

–

CM

–

–

–

–

–

–

SA

–

–

–

–

–

–

–

–

CS

–

–

–

–

– –

– –

– –

– –

– –

SA SA, NZ

– –

– –

– –

–

–

–

–

CM, SA

–

–

–

–

65

Sampling locality codes: AZ: Azores Islands; BE: Bering Sea; BR: Brazil Atlantic coast; BS: Barents Sea; CM: Central Mediterranean Sea; CS: Caribbean Sea; EM: East Mediterranean Sea; FA: Falkland Islands; FL: Florida coast; IC: Iberian Atlantic Coast; JA: Japan Sea; MA: Mauritanian coast; MD: Madeira Island; NAM: North African Mediterranean coast; NEA: NorthEast Atlantic; NEP: North-East Pacific; NWA: North-West Atlantic; NZ: New Zealand; PC: Portuguese coast; SA: South Africa coast; SC: Somali coast; SI: off Sakhalin Islands; TA: Tasman Sea; WM: West Mediterranean Sea (data from Mattiucci and Nascetti, 2006, 2007; Mattiucci et al., 1986, 1997, 2001, 2002, 2004, 2005; Nadler et al., 2005; Nascetti et al., 1986; Paggi et al., 1998a,b,c). Hosts listed by alphabetical order of the family.

66

TABLE 2.5

Intermediate/paratenic hosts so far detected, by molecular genetic markers, for the Anisakis spp.

Cephalopods Sepiidae Sepia officinalis Ommastrephidae Todaropsis eblanae Todarodes sagittatus Todarodes angolensis Illex coindettii Fishes Belonidae Belone belone Bothidae Arnoglossus laterna Arnoglossus imperialis Bramidae Brama brama Carangidae Trachurus capensis Trachurus mediterraneus Trachurus picturatus Trachurus trachurus

Selar crumenophthalmus

A. simplex (s.s.) A. pegreffii A. simplex C A. typica

Anisakis A. ziphidarum sp.

A. physeteris

A. brevispiculata A. paggiae

IC

–

–

–

–

–

–

–

–

IC, SA IC – IC

IC, SA NAM, CM SA –

– – – –

– – – –

– – – –

– – – –

– CM – –

– – – –

– – – –

IC

IC

–

–

–

–

–

–

–

PC –

– PC

– –

– –

– –

– –

– –

– –

– –

–

SA

–

–

–

–

–

–

–

– –

SA CM

– –

– –

– –

– –

– –

– –

– –

AZ, MD

AZ, MD

AZ, MD

–

NEA, IC, MA, WM –

CM, EM, WM, IC, MA –

–

EM

–

AZ

CM

–

–

–

CHS

–

–

–

–

–

Citharidae Citharus linguatula Clupeidae Clupea harengus

PC

NEA, BS, NEP Etrumeus whiteheadi – Congridae Conger conger IC Astroconger – myriaster Coryphaenidae Coryphaena – hippurus Emmelicththydae Emmelicththys – nitidus nitidus Engraulidae Engraulis – encrasicolus Gadidae Boreogadus saida IC Micromesistius IC poutassou Gadus morhua BS, IC Theragra NEP, JA, BE chalcogramma Trisopterus luscus IC Gempylidae Thyrsites atun –

PC

–

–

–

–

–

–

–

–

–

–

–

–

–

–

–

SA

–

–

–

–

–

–

–

CM CHS

– –

– –

– –

– –

– –

– –

– –

–

–

SC

–

–

–

–

–

SA

–

–

–

–

–

–

–

CM

–

–

–

–

–

–

–

– CM, IC, NAM – –

– –

– –

– –

– –

– –

– –

– –

– –

– –

– –

– –

– –

– –

– –

–

–

–

–

–

–

–

–

–

SA

–

–

–

–

–

–

67

(continued)

68

TABLE 2.5

(continued)

Hexagrammidae Pleurogrammus azonus Lophiidae Lophius piscatorius Lophius vomerinus Lotidae Molva dypterygia Brosme brosme Merluccidae Merluccius capensis Merluccius hubbsi Merluccius merluccius

Muraenidae Muraena helena Moridae Pseudophycis bachus Nemipteridae Nemipterus virgatus Nemipterus bathybius Ophidiidae Genypterus capensis

A. simplex (s.s.) A. pegreffii A. simplex C A. typica

Anisakis A. ziphidarum sp.

A. physeteris

A. brevispiculata A. paggiae

NEA

–

–

–

–

–

–

–

–

IC –

NAM SA

– –

– –

– –

– –

– –

– –

– –

IC NEA

– –

– –

– –

– –

– –

– –

– –

– –

– – NEA, IC, MA

SA FA CM, EM, WM, IC, NEA, MA

– – –

– – MA, EM, NAM

– – MA

– – IC

– – CM, MA, WM, IC, EM, NAM

– – MA

– – IC

–

NAM

–

–

–

–

–

–

–

–

NZ

NZ

–

–

–

–

–

–

– –

– –

– –

CHS CHS

– –

– –

– –

– –

– –

–

SA

–

–

–

–

–

–

–

Osmeridae Hypomesus pretiosus japonicus Phycidae Phycis phycis Phycis blennoides Pinguipedidae Parapercis colias Pleuronectidae Hippoglossus hippoglossus Platichthys flesus Salmonidae Oncorhynchus gorbuscha Oncorhynchus keta Salmo salar Scophtalmidae Scomberesocidae Scomberesox saurus Scombridae Lepidorhombus boscii Scomber japonicus Scomber scombrus

JA

–

–

–

–

–

–

–

–

– –

NAM –

– –

NAM –

– –

– –

NAM NAM

– –

– –

–

NZ

NZ

–

–

–

–

–

–

BE

–

–

–

–

–

–

–

–

–

–

–

PC

–

–

–

–

–

SI

–

–

–

–

–

–

–

–

SI NWA

– –

– –

– –

– –

– –

– –

– –

– –

NWA

–

–

–

–

–

–

–

–

IC

IC, NAM

–

–

–

–

–

–

–

AZ, MD NAM

AZ, MD –

– –

MD NAM

– –

– –

AZ, MD, JA AZ, MD, JA – NEA, IC, CM, IC – NAM

(continued)

69

70

TABLE 2.5

(continued)

Thunnus thynnus Auxis thazard Euthynnus affinis Sarda orientalis Scomberomorus commerson Scorpaenidae Scorpaena scrofa Sebastidae Helicolenus dactylopterus Soleidae Dicologlossa cuneata Solea senegalensis Sparidae Spondyliosoma cantharus Sternoptychidae Maurolicus muelleri Trachichthyidae Hoplostethus atlanticus Hoplostethus mediterraneus Trachinidae Echiichthys vipera Trichiuridae Lepidopus caudatus

A. simplex (s.s.) A. pegreffii A. simplex C A. typica

Anisakis A. ziphidarum sp.

A. physeteris

A. brevispiculata A. paggiae

JA – – – –

CM, BR – – – –

– – – – –

BR BR SC SC SC

– – – – –

– – – – –

– – – – –

– – – – –

– – – – –

IC

IC

–

–

–

–

–

–

–

–

CM, SA

–

–

–

–

–

–

–

– PC

PC –

– –

– –

– –

– –

– –

– –

– –

IC

–

–

–

–

–

–

–

–

NEA

–

–

–

–

–

–

–

–

–

–

TA

–

–

–

–

–

–

–

CM

–

–

–

–

–

–

–

–

NAM

–

–

–

–

–

–

–

–

CM, SA

–

–

–

–

–

–

–

Aphanopus carbo Trichiurus lepturus Triglidae Eutrigla gurnardus Xiphiidae Xiphias gladius

MD –

MD NAM

–

–

MD

AZ

AZ

AZ

AZ

IC

–

–

–

–

–

–

–

–

NEA

CM

–

CA

CA

–

CM, IC, NEA, CA, EM

AZ, CA

AZ

Sampling locality codes: AZ: Azores Islands; BE: Bering Sea; BR: Brazil Atlantic coast; BS: Barents Sea; CHS: China Sea; CA: Central Atlantic Ocean; CM: Central Mediterranean Sea; CS: Caribbean Sea; EM: East Mediterranean Sea; FA: Falkland Islands; FL: Florida coast; IC: Iberian Atlantic Coast; JA: Japan Sea; MA: Mauritanian coast; MD: Madeira Island; NAM: North African Mediterranean coast; NEA: North-East Atlantic; NEP: North-East Pacific; NWA: North-West Atlantic; NZ: New Zealand; PC: Portuguese coast; SA: South Africa coast; SC: Somali coast; SI: off Sakhalin Islands; TA: Tasman Sea; WM: West Mediterranean Sea (data from: Abollo et al., 2001; Farjallah et al., 2008; Klimpel et al., 2007; Marques et al., 2006; Mattiucci and Nascetti, 2006, 2007; Mattiucci et al., 1986, 1997, 2001, 2002a, 2004, 2005, 2007a; Nascetti et al., 1986; Orecchia et al., 1986a; Paggi et al., 1998a, b, c; Pontes et al., 2005; Umehara et al., 2006, 2008; Zhu et al., 2007). Hosts listed by alphabetical order of the family.

71

72

Simonetta Mattiucci and Giuseppe Nascetti

A. pegreffii Campana-Rouget and Biocca, 1955: Previously indicated as A. simplex A (see Nascetti et al., 1986), A. pegreffii is the dominant species of Anisakis in the Mediterranean Sea, being widespread in all the fish species. Indeed, it is presently the most important anisakid nematode in several pelagic and demersal fish from Mediterranean waters (Farjallah et al., 2008; Mattiucci et al., 1997, 2007a; Paggi et al., 1998a). It is also widely distributed at both adult and larval stage in the Austral Region between 30 N and 55 S (Mattiucci et al., 1997). In Atlantic waters, the northerly limit of its geographical range is represented by the Iberian coast (Abollo et al., 2001; Marquez et al., 2006; Mattiucci et al., 1997, 2004, 2007a; Pontes et al., 2005), and has not so far been reported from the western Atlantic (our unpublished data). It has been detected, using the PCR-RFLP of the ITS region, at the larval stage in some fish hosts from Japanese waters (Abe et al., 2006; Umehara et al., 2006, 2008) (Table 2.5; Fig. 2.1). SSCPbased identification of A. pegreffii larvae in fish from China waters, using genetic markers in the ITS-2 rDNA was also reported (Zhu et al., 2007) (Table 2.5; Fig. 2.1). To date, it has been recorded as an adult in three species of dolphins, belonging to the family Delphinidae, and in several species of fish and three squids as a larva (Tables 2.4 and 2.5). Among these, two definitive and sixteen intermediate/paratenic hosts were found to be shared with A. simplex (s.s.) in the contact area between the two species [Iberian Atlantic coast, western Mediterranean Sea (Alboran Sea) and Japan Sea waters] (Fig. 2.1). Whereas, two definitive and few intermediate/paratenic hosts are shared by A. pegreffii and A. simplex C in the Austral waters off New Zealand, the South African coast and the southern Chilean coast (Table 2.5; Fig. 2.1). Based on PCR-RFLP of the ITS, D’Amelio et al. (1999) and direct sequencing of the mtDNA cox2 (Mattiucci et al., 2007b), Anisakis larvae removed by endoscopy from humans in Italy were recognized as belonging to A. pegreffii. A. simplex C of Mattiucci et al. (1997): A. simplex C currently exhibits a discontinuous range, including the Canadian and Chilean Pacific coasts, New Zealand waters and the South African Atlantic coast (Mattiucci et al., 1997, unpublished data; Nadler et al., 2005). This species has been identified at the adult stage in cetaceans and as a larva it occurs syntopically with A. pegreffii in some fish species (Tables 2.4 and 2.5; Fig. 2.1). It has been occasionally identified also in Mirounga leonina from sub-Antarctic area (our unpublished data) and in M. angustirostris from North-East Pacific Ocean (Nadler et al., 2005). A. typica (Diesing, 1860): According to the data from the A. typica populations so far detected genetically, its range extends from 30 S to 35 N in warmer temperate and tropical waters (Tables 2.4 and 2.5; Fig. 2.1) (Mattiucci et al., 2002a). In these areas it was found by genetic markers as an adult in dolphin species and as a larva in several fish

Anisakid Nematodes and Host–Parasite Co-evolutionary Processes

73

species (Tables 2.4 and 2.5). A. typica has also been identified in the striped dolphin, Stenella coeruleoalba, and in the European hake, Merluccius merluccius, from the eastern Mediterranean Sea (off Cyprus). Its presence in these waters could be the result of the ‘Lesseptian migration’ (through the Suez Canal) (Mattiucci et al., 2004) of its intermediate/paratenic hosts from the Indian Ocean. It was recently recognized using RFLPPCR of the ITS rDNA in the flatfish Platichthys flesus captured in central Portuguese waters of the NE Atlantic Ocean (Marques et al., 2006), and only rarely in some fish caught along the North African coast of the Mediterranean Sea (Farjallah et al., 2008). SSCP-based identification of Anisakis spp. larvae, using genetic markers in the ITS-2 rDNA, detected A. typica larvae in fish from China waters (Zhu et al., 2007) (Table 2.5; Fig. 2.1). Anisakis sp. 1: Recently, a new gene pool, referred to as Anisakis sp. 1 was genetically detected by both allozyme and mtDNA analysis, at the larval stage, as a parasite of the fish Nemipterus japonicus caught off the Malaysian coast (Berland and Mattiucci, unpublished data). This new taxon is genetically distinct from all the known species of Anisakis but most closely related to A. typica from central Atlantic waters. Although known only at the larval stage, the third-stage larva of this undescribed taxon is a type I larva (sensu Berland, 1961). The preliminary results appear to indicate that this taxon may be a sibling species of A. typica occurring in central Pacific waters (Berland and Mattiucci, unpublished data). A. ziphidarum Paggi, Nascetti, Webb, Mattiucci, Cianchi and Bullini, 1998: A. ziphidarum was first described, both genetically and morphologically, as an adult in the beaked whales Mesoplodon layardii and Ziphius cavirostris from the South Atlantic Ocean (off the South African coast). Subsequently, it has also been recorded in the Mediterranean Sea, also in Z. cavirostris. Since its first morphological description and genetic characterization (Paggi et al., 1998b), it has recently been identified genetically as an adult in other species of beaked whale, such as M. mirus and M. grayi, in South Atlantic waters and in Mesoplodon sp. and Z. cavirostris in Caribbean waters (Mattiucci and Nascetti, 2006). Thus, its geographical range appears to be wide (Fig. 2.1) and related to that of its definitive hosts. Only scanty data are available concerning its infection in fish and/or squid, but it is responsible for a low prevalence of infection in some fish species, such as in Merluccius merluccius (see Mattiucci et al., 2004) and Aphanopus carbo (see Pontes et al., 2005; Saraiva et al., 2007) in central Atlantic waters. However, it seems that this species may involve other intermediate hosts, such as squid, rather than fish in its life cycle, as these represent the main food source of beaked whales. Anisakis sp. of Valentini et al. (2006): This species has been detected only at a larval stage (L4) in the beaked whales Mesoplodon mirus and M. grayi from South African and New Zealand waters (Mattiucci and

74

Simonetta Mattiucci and Giuseppe Nascetti

Nascetti, 2006) (Fig. 2.1; Table 2.4). The gene pool was found to be reproductively isolated from the sympatric species A. ziphidarum occurring in the same hosts and geographical region. It is genetically very distinct from the other species of Anisakis, but is most closely related to A. ziphidarum (see Section 5.1; Table 2.8). The third-stage larva of this undescribed taxon is apparently of type I and has on rare occasions been identified in the fishes (see Mattiucci et al., 2007a; Saraiva et al., 2007) caught in North East Atlantic waters. Conversely, this species has been genetically identified, at the larval stage, heavily infecting the squid Moroteuthis ingens in Tasman Sea waters (our unpublished data). This appears to support the hypothesis that this species involves squid rather than fish in its life cycle. Recently, Pontes et al. (2005) detected the occurrence of a new taxon at larval stage in Aphanopus carbo and Scomber japonicus from Madeira waters (Atlantic Ocean) indicated as Anisakis sp. A, genetically closely related to A. ziphidarum. However, the exact correspondence between this taxon and Anisakis sp. of Valentini et al. (2006) has not yet been assessed.

3.1.2. Anisakis spp. included in clade II Three species and one new gene pool of Anisakis so far comprises this clade, as clearly demonstrated by allozymes (Mattiucci et al., 2005) and mtDNA cox2 sequence analysis (Valentini et al., 2006). They are A. physeteris, A. brevispiculata and A. paggiae, which represent a complex of sibling species readily recognized genetically at both nuclear and mitochondrial levels. The existence of A. physeteris clustering with A. brevispiculata in the same clade, as a sister group to the other species included in the clade I, was also supported by rDNA ITS sequence phylogenetic analysis (Nadler et al., 2005) (see also Section 4.1). As far as is known, all these species share at the third larval stage the morphology known as type II (sensu Berland, 1961). A summary of each follows, including a synopsis of some ecological aspects of each species in relation to their known geographical distribution and both definitive and intermediate/paratenic hosts (Tables 2.4 and 2.5). A. physeteris (Baylis, 1920): A population of this species was first genetically characterized in its main definitive host, the sperm whale, Physeter macrocephalus, from Mediterranean waters (Mattiucci et al., 1986); no genetically identified adults have been recorded in other cetacean hosts. Type II larvae of A. physeteris have been genetically identified, occurring rarely in only a very few fish host species (Mattiucci et al., 1986, 2001, 2004), except for the swordfish, Xiphias gladius, from Mediterranean and Atlantic waters in which it represents the main Anisakis species (Mattiucci et al., 2007a) (Tables 2.4 and 2.5; Fig. 2.1). Despite the fact that the swordfish might only represent an accidental host in the life cycle of

Anisakid Nematodes and Host–Parasite Co-evolutionary Processes

75

Anisakis spp., acquiring the infection by preying on infected invertebrates or squid (main prey of the swordfish), suggests that other intermediate hosts, mainly squids rather than fish are involved in the life cycle of this parasite. A. brevispiculata Dollfus, 1966: A population of A. brevispiculata was initially characterized genetically using allozymes based on material from a pygmy sperm whale, Kogia breviceps, in South African and North East Atlantic waters (Iberian coast) (Mattiucci et al., 2001) (Table 2.4; Fig. 2.1). Reproductive isolation from the morphologically closely related A. physeteris was demonstrated, establishing the validity of A. brevispiculata (see Mattiucci et al., 2001), which has been synonymized with A. physeteris by Davey (1971). Later, the same species was sequenced at the mtDNA cox2 gene and its genetic relationships with respect to the other Anisakis spp. was established, confirming that A. brevispiculata clusters well with those Anisakis species forming the second clade, as also indicated by nuclear markers (allozymes). A congruent result was inferred from the ITS rDNA sequence analysis (Nadler et al., 2005) (see Section 4.1). Anisakis larvae of type II corresponding to A. brevispiculata were recognized by allozyme markers as a rare parasites of the fish Merluccius merluccius (see Mattiucci et al., 2004, 2007a) and the swordfish Xiphias gladius in Atlantic waters (Mattiucci et al., 2007a). A. paggiae Mattiucci et al. (2005): In the cluster formed by A. physeteris and A. brevispiculata, this third species has been recently demonstrated by both allozymes (Mattiucci et al., 2005) and mtDNA cox2 sequence analysis (Valentini et al., 2006) (Fig. 2.1). A. paggiae was first genetically characterized and described morphologically as an adult parasite of the pygmy sperm whale, Kogia breviceps, and the dwarf sperm whale, K. sima, from off both Florida and the South African Atlantic coast (Mattiucci et al., 2005). Scanty data are so far available regarding the identification of the intermediate hosts in the life cycle of A. paggiae. A very few larvae of type II have been identified as belonging to these species in fish from Atlantic waters (i.e. M. merluccius and X. gladius) (see Mattiucci et al., 2007a, unpublished data), thus suggesting that other hosts, not yet detected, are involved in the life cycle of this Anisakis species. Anisakis sp. 2: A further gene pool, referred to as Anisakis sp. 2, has been detected genetically by means of allozyme markers and mtDNA cox2 sequence analysis based on larvae of type II from the swordfish X. gladius in the equatorial area (Mattiucci et al., 2007a, unpublished data). The new taxon was shown to be genetically distinct from all the other Anisakis species, and that it is mostly closely related to A. physeteris. Preliminary phylogenetic analysis showed that it clusters with the clade formed by A. physeteris, A. brevispiculata and A. paggiae, thus suggesting that this new gene pool might represent an undescribed sibling species belonging to this same complex.

76

Simonetta Mattiucci and Giuseppe Nascetti

The high genetic heterogeneity of the Anisakis spp. is now also supported by some differential morphological features detected in the species belonging to this genus, where the two major clades can be delineated as follows: (i) the ventriculus, in the adult stage, is short, never sigmoid and broader than long in A. physeteris, A. brevispiculata and A. paggiae (see Mattiucci et al., 2005), and longer than broad and often sigmoid in shape in those species included in clade I (the species of the A. simplex complex, A. typica and A. ziphidarum) (see Mattiucci et al., 2005; Paggi et al., 1998b); (ii) male spicules that are short, stout and of similar length can be observed in A. physeteris, A. brevispiculata and A. paggiae (Mattiucci et al., 2005) but are long and often unequal (equal in A. ziphidarum; see Paggi et al., 1998b) in species of clade I; and (iii) type II larval morphology (sensu Berland, 1961) is characteristic of A. physeteris, A. brevispiculata, A. paggiae and Anisakis sp. 2 (Mattiucci et al., 2001, 2004, 2005, 2007a) (clade II), whereas a type I morphology (sensu Berland, 1961), can be found in all the species of the A. simplex complex, A. typica, A. ziphidarum and Anisakis sp. and Anisakis sp. 1 (clade I). While no morphological characters are so far known which help in distinguishing the sibling species of the A. simplex complex, some morphological features, of diagnostic value, available in male and female adult specimens, were used to help in distinguishing A. paggiae from A. physeteris and A. brevispiculata (see Mattiucci et al., 2005). Indeed a morphological key to the recognized adults of Anisakis spp. so far included in clade II was provided by Mattiucci et al. (2005).

3.2. The current taxonomy of Pseudoterranova decipiens (sensu lato) The species decipiens was first described as Ascaris decipiens by Krabbe (1878). A. decipiens was later linked with Terranova Leiper and Atkinson, 1914 (a genus erected for parasites of elasmobranchs) by Baylis (1916), because of the presence of an intestinal caecum. Subsequently, Baylis linked this species within Porrocaecum Railliet and Henry, 1912 (a group of ascarids now mainly restricted to terrestrial birds), because he considered it as a senior synonym of Terranova. Terranova was resurrected at full generic level by Johnston and Mawson (1945) and as a subgenus by Karokhin (1946), the latter being the first to use the combination T. decipiens. T. decipiens was accepted by Mozgovoi (1951, 1953), Hartwich (1957) and Yamaguti (1961). However, because Terranova included species maturing in different definitive hosts (elasmobranchs, reptiles, marine mammals), Myers (1959) erected Phocanema (with decipiens as the type and only species) for those species included in Terranova which were parasites of marine mammals on the base of some morphological features at the cephalic end (structure of the labia) and the male caudal end.

Anisakid Nematodes and Host–Parasite Co-evolutionary Processes

77

Accordingly, the species decipiens and Terranova kogiae Johnston and Mawson, 1945 (described from the pygmy sperm whale, Kogia breviceps) were included in Phocanema. Later, Gibson (1983) considered Phocanema as a synonym of Pseudoterranova, previously erected by Mozgovoi (1951; listed in Mozgovoi, without indication) for parasites from sperm whales, and decipiens was transferred to this genus (combination first used in Gibson and Colin, 1982). According to the same authors, the group of species of Pseudoterranova from marine mammals included P. kogiae (Johnston and Mawson, 1945), P. ceticola (Deardoff and Overstreet, 1981) and P. decipiens. This is the status quo which has been widely accepted. Although several nominal species related to Pseudoterranova decipiens (i.e. occurring in seals and possessing only an intestinal caecum) having been described, these species have tended to be considered as synonyms of P. decipiens. This species appeared to have a low host specificity, with up to 18 seal species been recorded as definitive hosts (e.g. Hartwich, 1975; King, 1983; Myers, 1959), and exhibited some morphological variation between specimens collected from the bearded seal, Erignathus barbatus, in Canadian waters (Baylis, 1916). Population genetic analysis, first performed by allozyme markers, on specimens of P. decipiens (s.l.) recovered from 10 fish species, and from four seal species, collected at several locations of the North Atlantic Ocean from the Canadian Atlantic eastwards to the Barents Sea, demonstrated the existence of a remarkable genetic heterogeneity with striking variation in allele frequencies among the samples (Paggi et al., 1991). Indeed, the observed genotype frequencies at various loci showed significantly deviations from those expected using the Hardy–Weinberg equilibrium, with a complete absence or marked deficiency of various heterozygous classes in material recovered from several sites and also between worms recovered from one host individual. The only possible explanation was that three distinct biological species occurred sympatrically in the samples of P. decipiens (s.l.) collected in seals hosts from those geographical areas, with no gene flow between them: the three taxa genetically recognized were thus provisionally designated as P. decipiens A, P. decipiens B and P. decipiens C (see Paggi et al., 1991). Morphological analysis carried out on male specimens identified by allozyme markers as P. decipiens A and B allowed the detection of significant differences in a number of characters between these two members; on the basis of such differences the nomenclature designation for P. decipiens A and P. decipiens B was proposed (see Paggi et al., 2000). The names Pseudoterranova krabbei Paggi et al., 2000 and P. decipiens (s.s.) were proposed, respectively, for species A and B, and a formal description of the two taxa was provided (see Paggi et al., 2000). The name of P. bulbosa (Cobb, 1888) was proposed for the taxon P. decipiens C (see Mattiucci et al., 1998), as the latter taxon was demonstrated

78

Simonetta Mattiucci and Giuseppe Nascetti

to correspond morphologically with Ascaris bulbosa described by Cobb (1888) from the collection of Dr. Kukenthal collected from the bearded seal, Erignathus barbatus, at Spitzbergen (NE Atlantic Ocean). Furthermore, a sample of P. decipiens C from that same locality was studied both genetically and morphologically by Paggi et al. (1991, 2000). Previously, Ascaris bulbosa had been considered a synonym of Phocanema decipiens [later included in Pseudoterranova by Gibson and Colin (1982)] by Myers (1959). Therefore, Mattiucci et al. (1998) proposed the name P. bulbosa (Cobb, 1888) n. comb. for P. decipiens C. A further taxon, provisionally designated as P. decipiens D (see Mattiucci et al., 1998), was later included in the P. decipiens complex; this was detected by exhibiting several fixed differences at some enzyme loci with respect to P. decipiens A, P. decipiens B and P. decipiens C. It was found to occur sympatrically with P. bulbosa in the same geographical areas (Japanese waters) and occasionally in the same definitive host, the bearded seal Erignathus barbatus, from which it was demonstrated to be reproductively isolated (Mattiucci et al., 1998). P. decipiens D was found to correspond to the measurements and tail drawing of Porrocaecum azarasi Yamaguti and Arima, 1942, a species described by Yamaguti and Arima (1942) based on specimens recovered in the ribbon seal Phoca (=Histriophoca) fasciata on the islands of Sakhalin and Hokkaido. This taxon was synonymized by Margolis (1956) with ‘Phocanema decipiens’. Therefore, Mattiucci et al. (1998) proposed the name Pseudoterranova azarasi (Yamaguti and Arima, 1942) n. comb. for species P. decipiens D. Using allozyme markers on larval and adult population of P. decipiens (s.l.) collected from four fish species and the Austral fur seal, Otaria byronia, in the SE Pacific Ocean, a further member of the P. decipiens complex has been shown to exist (George-Nascimento and Llanos, 1995). In the formal description, this taxon was named P. cattani GeorgeNascimento and Urrutia, 2000. A summary of the sibling species P. decipiens (s.l.) recognized by using several molecular genetic methodologies are reported below with ecological data on their host preference (Table 2.5) and geographical range (Fig. 2.1). P. decipiens (Krabbe, 1868) (sensu stricto) (=P. decipiens B): This species was initially recognized as P. decipiens B by Paggi et al. (1991); its formal description was later given based on morphological features considered to be of diagnostic value with respect to the other members included in the P. decipiens complex (see Paggi et al., 2000). Its geographical range appears to be wide and mainly in the Arctic and sub-Arctic regions, including the North East Atlantic (comprising Scottish, Faroes, southern Icelandic and Norwegian waters), the Canadian Atlantic (including Newfoundland waters and the Gulf of S. Lawrence) (Brattey and Stenson, 1993; Paggi et al., 1991) and the Canadian Pacific waters (Mattiucci et al.,

Anisakid Nematodes and Host–Parasite Co-evolutionary Processes

79

1998; Nadler et al., 2005; Paggi et al., 1998c). Areas of sympatry between P. decipiens (s.s.) and other species of the P. decipiens complex have been detected; in the North Eastern Atlantic waters in mixed infections with both P. krabbei and P. bulbosa and in Canadian Atlantic waters with P. bulbosa (see Paggi et al., 1991) (Fig. 2.1). P. decipiens (s.s.) was recognized, using genetic markers, as a parasites at the adult stage in the common seal, Phoca vitulina, grey seal, Halichoerus grypus, in both North and West Atlantic waters, and in Phoca vitulina richardsii and Zalophus californianus in northern Pacific waters (Paggi et al., 1998c). Its larval forms have been identified in some gadoid fish species of the North Atlantic Ocean (Mattiucci et al., 1998; Paggi et al., 1991) (Table 2.6). P. krabbei Paggi, Mattiucci et al., 2000: This species was previously referred to as P. decipiens A (Paggi et al., 1991). To date it has been found only in the North-East Atlantic (including off Scotland, the Faroes, southern Iceland and Norway) (Fig. 2.1). It is mainly an adult parasite of the grey seal, Halichoerus grypus, and the common seal, Phoca vitulina, often in mixed infections in the same individual host with P. decipiens (s.s.) (Table 2.6). However, in the eastern Atlantic waters, when detected in sympatry with P. decipiens (s.s.), P. krabbei prevails in the grey seal rather than in the common seal. The ecological significance of this finding is discussed in Section 7. P. krabbei has been recognized genetically at the larval stage as a parasite of Gadus morhua, Melanogrammus aeglefinus and Pollachius virens (see Paggi et al., 1991) (Table 2.6; Fig. 2.1). P. krabbei exhibits, in male specimens with respect to P. decipiens (s.s.), the following diagnostic morphological features: shorter spicules; a proximal papilla (nomenclature according to Fagerhom, 1989) smaller than d1 versus the same papilla size in P. decipiens (s.s.); distal papillae 1, 2 and 4 closer to each other; and caudal plates of a similar width and narrower than in P. decipiens (s.s.) where plates (wp) 1 (wp1) and 2 (wp2) are of a similar width but plate 3 (wp3) is narrower (Paggi et al., 2000). P. bulbosa (Cobb, 1888): Previously referred to as P. decipiens C (Paggi et al., 1991) (see Mattiucci et al., 1998), this species has been recorded from the Barents and Norwegian Seas, the Canadian Atlantic and the Sea of Japan, between 40 N and 80 N (Brattey and Stenson, 1993; Mattiucci et al., 1998; Paggi et al., 1991). Its main definitive host so far detected is the bearded seal, Erignathus barbatus. In this phocid host from the Otaru Sea (Sea of Japan), P. bulbosa has been identified as occurring in mixed infections with P. azarasi (see Mattiucci et al., 1998). The benthic flatfishes Hippoglossoides platessoides and Reinhardtius hippoglossoides are reported as its intermediate hosts (Paggi et al., 1991) (Table 2.6; Fig. 2.1). P. bulbosa differs from the other members of the P. decipiens complex in having the following morphological features in adult male specimens: longer spicules; caudal plates (wp) of unequal width, with plate 2 (wp2)

80 TABLE 2.6 Definitive and intermediate/paratenic hosts so far detected, by molecular genetic markers, for the Pseudoterranova decipiens species complex

Pinnipeds Otariidae Eumetopias jubatus Otaria byronia Zalophus californianus Phocidae Phoca vitulina richardsii Phoca vitulina Erignathus barbatus Halichoerus grypus Cystophora cristata Mirounga angustirostris Leptonychotes weddellii Fishes Channichthydae Chaenocephalus aceratus

P. krabbei

P. decipiens (s.s.)

P. bulbosa

P. azarasi

P. decipiens E

P. cattani

–

–

–

–

–

– –

– NEP

– –

NWP, JA – NWA

– –

SEP –

– NEA

– –

NWA –

– –

– –

–

NEP NEA, NWA, NEP –

–

NEA, NWA NWA, NEA NEP –

NWP, JA – – – –

–

NEA – – –

NEP, NEA, LS – – – –

– – – AN

– – – –

–

–

–

–

AN

– (continued)

TABLE 2.6 (continued)

Cottidae Myxocephalus scorpius Myxocephalus quadricornis Gadidae Gadus morhua macrocephalus Gadus morhua Gadus ogac Pollachius virens Melanogrammus aeglefinus Boreogadus saida Lotidae Brosme brosme Merluccidae Merluccius gayi Ophidiidae Genypterus maculatus Osmeridae Osmerus eperlanus

P. krabbei

P. decipiens (s.s.)

P. bulbosa

P. azarasi

P. decipiens E

P. cattani

– –

NWA –

– NWP

– BE, NWP

– –

– –

–

NWA

NWP

–

–

NEA – FI FI

NEA NWA NWA –

– – – –

JA, NWP – – – –

– – – –

– – – –

–

NWA

–

–

–

–

–

NEA

–

–

–

–

–

–

–

–

–

SEP

–

–

–

–

–

SEP

–

NEA

–

–

–

–

81

(continued)

TABLE 2.6 (continued)

Notothenidae Notothenia coriiceps Notothenia neglecta Trematomus newnesi Paralichthydae Paralichthys microps Pleuronectidae Hippoglossus hippoglossus Hippoglossoides platessoides Reinhardtius hippoglossoides Scophthalmidae Psetta maxima

P. krabbei

P. decipiens (s.s.)

P. bulbosa

P. azarasi

P. decipiens E

P. cattani

– – –

– – –

– – –

– – –

AN AN AN

– – –

–

–

–

–

–

SEP

–

–

NWP

–

–

NEA

NEA

BS

BE, NWP –

–

–

–

–

NWA

–

–

–

NEA

–

–

–

–

–

Codes: AN: Antarctica; BE: Bering Sea; BS: Barents Sea; FI: Faeroe Islands; JA: Japan Sea; LS: Labrador Sea; NEA: North-East Atlantic; NEP: North-East Pacific; NWA: NorthWest Atlantic; NWP: North-West Pacific; SEP: South-East Pacific (Chilean coast) (data from George-Nascimento and Llanos, 1995; George-Nascimento and Urrutia, 2000; Mattiucci et al., 1998; Nadler et al., 2005; Paggi et al., 1991, 1998c; Zhu et al., 2002). Hosts listed by alphabetical order of the family.

Anisakid Nematodes and Host–Parasite Co-evolutionary Processes

83

narrower than 1(wp1) and 3 (wp3); and a different pattern of distal caudal papillae, with d2 far apart from d4 (Mattiucci et al., 1998). P. azarasi (Yamaguti and Arima, 1942): This species was previously referred to as P. decipiens D (see Mattiucci et al., 1998). Its main definitive host has been identified as the Steller’s sea lion, Eumetopias jubatus. Its geographical distribution appears limited to Japanese and Sakhalinese waters of the North Pacific Ocean (Mattiucci et al., 1998) (Table 2.6; Fig. 2.1). Both P. bulbosa and P. azarasi were resurrected, on both allozyme and morphological bases (see Mattiucci et al., 1998), from synonymy with P. decipiens (see Margolis, 1956; Myers, 1959). P. azarasi differs from the other members of the P. decipiens complex in having shorter spicules; caudal plates (wp) of increasing width from 1 to 3; and a different pattern of distal papillae (sensu Fagerholm, 1989), with d2 closer to d4 than in P. bulbosa but more distant than in P. krabbei and P. decipiens (s.s.) (see Mattiucci et al., 1998). P. decipiens E of Bullini et al., 1997. This taxon was genetically detected in the Antarctic Weddell seal, Leptonychotes weddellii (see Bullini et al., 1997). Scanty data are so far available in relation to the nature of its intermediate host, although larvae corresponding to this sibling species were found as parasites of benthic fish hosts in subantartic waters (see Section 7.2) (Table 2.6). P. cattani George-Nascimento and Urrutia, 2000: As stated above, this species was found as an adult in Otaria byronia on the Chilean coast (Table 2.5; Fig. 2.1). Using molecular markers in the internal transcribed spacers of ribosomal DNA (ITS-rDNA), this species was recently shown to cluster with the P. decipiens complex (Zhu et al., 2002) (Section 4.2).

3.3. The current taxonomy of Contracaecum species from pinnipeds Contracaecum was originally defined by Railliet and Henry (1912) based on the morphology of the oesophago-intestinal region. C. spiculigerum (Rudolphi, 1809) (=Ascaris spiculigerum, Rudolphi, 1809) was designated as the type-species. The original type-material was later identified as C. microcephalum (Rudolphi, 1809) by Hartwich (1964), and this species was accordingly defined as the new type-species of the genus. Presently, the genus comprises some 50 nominal species, most of which are adults in pinnipeds and fish-eating birds. Hartwich (1975) defined the genus based on morphological details; important features are the presence of interlabia and the position of the excretory pore at the base of the ventral interlabium. However, several later authors suggested that the genus should be divided or amended. Mozgovoi (1951) had previously divided the genus

84

Simonetta Mattiucci and Giuseppe Nascetti

into three subgenera: Contracaecum (Erschovicaecum) [clearly representing Hysterothylacium (see Deardoff and Overstreet, 1980) and including species maturing in fish], Contracaecum (Contracecum) and Contracaecum (Ornitocaecum). These two new subgenera were considered as synonyms of Contracaecum by Hartwich (1975). Phocascaris Ho¨st, 1932 included the following definition ‘interlabia present, reduced or absent . . . . parasites of the digestive tract of seals’, P. phocae being the type-species of the genus. According to that definition, Berland (1964), when describing P. cystophorae from the seal Cystophora cristata, suggested that the species of Contracaecum ‘with opposite caeca from seals’ should be transferred to the genus Phocascaris. His proposal was to retain within Contracaecum species from fish-eating birds. Berland’s (1964) hypothesis was mainly based on the life-cycle patterns of these anisakid nematodes and their definitive hosts, irrespective of their morphological phenotype. However, a taxonomic proposal based on those biological features was not generally accepted by taxonomists, and Hartwich (1975) did not take into account that observation when he produced the ‘Key of Ascaridoidea’, still considering a morphological character (relating to the presence/absence of interlabia) as a valid criterion for distinguishing the species belonging to Contracaecum and Phocascaris. Several years later, the same morphological character was shown to be inconsistent and not of useful systematic value for differentiating the two groups of species included in Contracaecum. Indeed, based on allozymes (22 enzyme loci), Orecchia et al. (1986b) and Nascetti et al. (1990) used species Contracaecum from seals and fish-eating birds to show that of all the species of the Contracaecum osculatum complex (see Section 4.3) are genetically much more closely related to species of Phocascaris (i.e. Phocascaris phocae and P. cystophorae) from seals than to the congeneric species from fish-eating birds, demonstrating high levels of differentiation (Contracaecum species from pinnipeds share no alleles in common with those from fish-eating birds). Clustering methods based on allozyme markers showed that P. phocae and P. cystophorae form a clade with the species of Contracaecum from seals (Nascetti et al., 1990) (Section 4.3), suggesting an evolutionary hypothesis for the systematic status of these species. Nadler et al. (2000), based on nuclear rDNA sequence data of several taxa of the genera Contracaecum and Phocascaris, demonstrated the validity of the evolutionary hypothesis previously suggested by allozyme markers. A phylogenetic hypotheses based on different clustering analyses of the ITS-rDNA (Nadler et al., 2000) and the mtDNA cytochromoxidase-2 (mtDNA cox2) (Mattiucci et al., 2008b) sequences data strongly supported the hypothesis based on allozymes, according to which species of Phocascaris are nested within the clade formed by the Contracaecum species hosted by phocid seals and are thus closely related to species of the Contracaecum osculatum complex. All of the phylogenetic analyses also support the hypothesis

Anisakid Nematodes and Host–Parasite Co-evolutionary Processes

85

advanced by Berland (1964), based on the biology of these anisakid nematodes, in terms of the monophyly of the species of Contracaecum and Phocascaris from phocids (see also Section 4.3). On the other hand, the molecular genetic approach has been demonstrated to be very fruitful in the detection of sibling species and disclosing new taxa of Contracaecum parasites of fish-eating birds (Bullini et al., 1986; D’Amelio et al., 1990, 2007; Li et al., 2005; Mattiucci et al., 2002b, 2008c), which are not however treated in details in this review.

3.3.1. The Contracaecum osculatum species complex To date, five members of the C. osculatum complex have been recognized genetically by the use of allozyme markers. They are the three Arctic sibling species referred to as C. osculatum A, C. osculatum B and C. osculatum (s.s.) (see Nascetti et al., 1993), and the two Antarctic members, C. osculatum D and C. osculatum E (see Orecchia et al., 1994). C. osculatum A of Nascetti et al. (1993): This species occurs in the Norwegian and Barents Seas, Canadian Atlantic, Icelandic, Canadian Pacific waters and the Sea of Japan, between 40 N and 80 N. Its main definitive host is the bearded seal, Erignathus barbatus, in both eastern and western part of the North Atlantic Ocean, but the grey seal, Halichoerus grypus, has also been recorded as a host (Brattey and Stenson, 1993; Nascetti et al., 1993). Subsequently, it has been detected genetically in the Steller’sea lion, Eumetopias jubatus, in the Western Pacific Ocean (Sea of Japan) (Mattiucci et al., 1998; Paggi et al., 1998c). Its larvae have been identified in the gadoid fish Theragra chalcogramma from the Sea of Japan (Mattiucci et al., 1998) (Table 2.7; Fig. 2.1). C. osculatum B of Nascetti et al. (1993): This species was first detected genetically as an adult in the phocid seals Pagophilus groenlandicus, Phoca vitulina and Halichoerus grypus from the North-eastern and north-western Atlantic Ocean (Brattey and Stenson, 1993; Nascetti et al., 1993); reproductive isolation was demonstrated by allozyme markers between this species and the other two Arctic members of the C. osculatum complex (Nascetti et al., 1993). Later, it was identified based on diagnostic allozyme markers as a parasite of the Phoca vitulina and Zalophus californianus in northern Pacific waters (Mattiucci et al., 1998) (Table 2.7; Fig. 2.1). C. osculatum (Rudolphi, 1802) (sensu stricto): This species, named C. osculatum C by Nascetti et al. (1993), was genetically characterized from the grey seal, Halichoerus grypus; its reproductive isolation from C. osculatum B was shown in the sympatric situation of individual seal hosts (i.e. the grey seal) from Iceland waters (Nascetti et al., 1993). It is the only species of the C. osculatum complex present in the Baltic Sea. So far, this species has not been reported from the western part of the Atlantic or from Pacific waters (Table 2.7; Fig. 2.1).

86

TABLE 2.7

Definitive and intermediate/paratenic hosts so far detected, by molecular genetic markers, for the Contracaecum spp. from pinnipeds

Pinnipeds Phocidae Phoca vitulina

C. osculatum A

C. osculatum B

C. osculatum (s.s.)

C. osculatum D

C. osculatum E

C. o. baicalensis

C. ogmorC. radiatum C. mirounga hini (s.s.)

C. margolisi

–

NWA, NEP, JA, BE NEA, NWA, JA, BS – BS

–

–

–

–

–

–

–

–

–

–

–

–

–

–

–

–

– –

– –

– –

BL –

– –

– –

– –

– –

Pagophilus groenlandicus

–

Phoca sibirica Erignathus barbatus

– NEP, NWA, NEA, JA, BS –

–

–

–

–

–

–

AN

–

–

–

NEP

–

–

–

–

–

–

–

–

NWA, NEA

NWA, NEA

–

–

–

–

–

–

–

–

–

BS, NEA, NWA, BA –

RS, WS

RS, WS

–

RS, WS

–

–

–

–

NWA

–

–

–

–

–

–

–

–

–

–

–

–

–

–

–

–

Mirounga leonina Eumetopias jubatus Halichoerus grypus Leptonychotes weddellii Cystophora cristata Otariidae Zalophus californianus

NEP

Arctocephalus australis Arctocephalus pusillus Fishes Bathydraconidae Gymnodraco acuticeps Cygnodraco mawsonii Channichthydae Cryodraco antarcticus Chionodraco hamatus Pagetopsis macropterus Chaenodraco wilsoni Cottidae

–

–

–

–

–

–

–

AR

AR

–

–

–

–

–

–

–

–

–

SA, AU

–

–

–

–

RS

RS

–

–

–

–

–

–

–

–

RS

RS

–

–

–

–

–

–

–

–

RS

RS

–

RS

–

–

–

–

–

–

RS

RS

–

RS

–

–

–

–

–

–

RS

RS

–

–

–

–

–

–

–

–

RS

RS

–

–

–

–

–

–

–

BS

–

Myoxocephalus quadricornis – Gadidae Gadus morhua macrocephalus Theragra chalcogramma Notothenidae

–

–

–

–

–

BE

–

–

–

–

–

–

–

–

–

BE

–

–

–

–

–

–

–

–

–

–

–

–

RS

RS

–

–

–

–

–

87 (continued)

TABLE 2.7

(continued) C. osculatum A

C. osculatum B

C. osculatum (s.s.)

C. osculatum D

C. osculatum E

C. o. baicalensis

C. ogmorC. radiatum C. mirounga hini (s.s.)

C. margolisi

–

–

–

RS

RS

–

–

–

–

–

–

–

–

RS

RS

–

–

–

–

–

BE

–

–

–

–

–

–

–

–

–

Notothenia neglecta Trematomus bernacchii Trematomus pennelli Pleuronectidae Hippoglossus hippoglossus

Sampling locality codes: AN: Antarctica; AR: Argentine coast; AU: Australian coast; BA: Baltic Sea; BE: Bering Sea; BL: Baikal Lake; BS: Barents Sea; JA: Japan Sea; NEA: North-East Atlantic; NEP: North-East Pacific; NWA: North-West Atlantic; NWP: North-West Pacific; RS: Ross Sea; SA: South Atlantic Ocean (off South Africa coast); SEA: South-East Atlantic; SWA: South-West Atlantic; WS: Weddell Sea (data from Mattiucci and Nascetti, 2007; Mattiucci et al., 1998, 2003, 2008b; Nadler et al., 2000, 2005; Nascetti et al., 1993; Orecchia et al., 1994; Paggi et al., 1998c). Hosts listed by alphabetical order of the family.

Anisakid Nematodes and Host–Parasite Co-evolutionary Processes

89

C. osculatum D and C. osculatum E of Orecchia et al. (1994): The Antarctic members of the C. osculatum complex, C. osculatum D and C. osculatum E were demonstrated to be reproductively isolated by allozyme markers. They occur sympatrically in the same definitive host, the Weddell seal, Leptonychotes weddellii, and have so far been reported from both the Weddell and the Ross Seas (Antarctica) (Orecchia et al., 1994). The larval stages of the two sibling species have been identified by allozyme markers from several fish species belonging to the families Channicthydae, Bathydraconidae and Nototheniidae, in which a differential distribution of the two sibling species is reported (Mattiucci and Nascetti, 2007; see also Section 7.2) (Table 2.7; Fig. 2.1).

3.3.2. The Contracaecum ogmorhini species complex The pinniped parasite Contracaecum ogmorhini Johnston and Mawson, 1941, first described from the leopard seal, Hydrurga leptonyx, in South Australian waters, was later synonymized with C. osculatum (see Johnston and Mawson, 1945). However, it was considered valid by Fagerholm and Gibson (1987). The species was found to be genetically heterogenous using allozyme markers (18 enzyme loci), indicating the existence of two reproductively isolated taxa (sibling species) included within the morphospecies. A formal description of the two taxa was given by Mattiucci et al. (2003), and they were named C. ogmorhini Johnston and Mawson (1941) (sensu stricto) and C. margolisi Mattiucci et al. (2003). A morphological description of C. ogmorhini (s.s.) from Arctocephalus australis was given by Timi et al. (2003). C. ogmorhini (s.s.) has been detected as an adult in the otariid seals Arctocephalus pusillus pusillus, A. pusillus doriferus and A. australis in the Austral region, while C. margolisi was detected in the otariid Zalophus californianus in the Boreal region (Table 2.6; Fig. 2.1) (see also Section 4.3). To date, the allopatric distribution of these two sibling species appears to be related to that of their definitive hosts, ranging from 20 to 55 N in Boreal Pacific waters, but from 20 to 50 S in South Atlantic and South Pacific waters.

3.3.3. Validity of species of Contracaecum, parasitic in seals, using molecular markers C. osculatum baicalensis Moszgovoi and Ryzhykov, 1950: The specific status of C. osculatum baicalensis was established genetically using allozyme markers (22 enzyme loci) with respect to the species of the C. osculatum complex listed above (D’Amelio et al., 1995). It was found to be genetically well distinct from the other members of this complex, as well as from the other congeneric taxa (Mattiucci et al., 2008b). C. osculatum baicalensis is an adult parasite of the freshwater Baikal seal, Phoca sibirica, endemic to Lake Baikal. The genetic relationships between C. osculatum baicalensis and other congeneric taxa were inferred from

90

Simonetta Mattiucci and Giuseppe Nascetti

LSU rDNA sequences (Nadler et al., 2000) and from the mtDNA cox2 sequence (Mattiucci et al., 2008b) analyses. Both phylogenetic analyses were congruent with the allozyme clustering in showing C. osculatum baicalensis nested within the clade formed by the species of the C. osculatum complex (see Section 4.3). The genetic relationships found between C. osculatum baicalensis and the members of the C. osculatum complex suggest that the evolutionary divergence of this taxon likely took place following its host’s isolation during a Pliocene-Pleistocene refuge (Deme´re´ et al., 2003). At a larval stage, C. osculatum baicalensis was recognized by allozymes in the endemic fish species Cottomephorus grewingki, C. inermis, Comephorus baicalensis, Coregonus lavaretus, C. autumnalis migratorius and Thymallus arcticus, which represent prey items of Phoca sibirica in Lake Baikal (our unpublished data). C. radiatum (v. Linstow, 1907) Baylis, 1920: The taxonomic status of this species was confirmed genetically by Arduino et al. (1995) on the basis of 24 enzyme loci. Several allozymes were found to be diagnostic between C. radiatum and the other taxa so far characterized as belonging to Contracaecum species from seals (Arduino et al., 1995; Mattiucci et al., 2008b, unpublished data). Reproductive isolation from the two Antarctic members of the C. osculatum complex (i.e. C. osculatum D and C. osculatum E) occurring sympatrically in the same definitive hosts (the Weddell seal) was proved by the lack of F1 hybrids and recombinant or introgressed individuals between the Antarctic taxa in the sympatric areas of the Weddell and Ross Seas (Arduino et al., 1995). The genetic relationships between C. radiatum and other congeneric taxa were later inferred from LSU rDNA sequences (Nadler et al., 2000) and mtDNA cox2 sequence analyses (Mattiucci et al., 2008b). Morphological distinction between C. radiatum and C. osculatum (s.l.) was given by Klo¨ser and Plo¨tz (1992). C. radiatum has been genetically identified as an adult in Leptonychotes weddellii and as a larva in the pelagic channichthyd fishes Chionodraco hamatus and Criodraco antarcticus (see Arduino et al., 1995). This finding supports a previous report by Klo¨ser et al. (1992), according to which C. radiatum has become adapted to a pelagic food web. Other definitive hosts recorded for this species in Antarctic waters are the leopard seal, Hydrurga leptonyx, and the Ross seal, Ommatophoca rossi (see Baylis, 1937; Dailey, 1975). Genetic investigations on this parasite of Antarctic seals are needed in order to determine any host preference of this species. C. mirounga Nikolskii, 1974: The taxonomic status of the species was confirmed genetically by allozyme markers (20 enzyme loci) (Mattiucci et al., 2008b, unpublished data). It was detected genetically as an adult in Mirounga leonina from the Antartic and sub-Antarctic area (Mattiucci et al., 2008b) and also in the otariid Arctocephalus australis (see Mattiucci et al., 2003). Several allozymes were found to be diagnostic between C. mirounga and other Contracaecum species from seals (Mattiucci et al., 2008b,

Anisakid Nematodes and Host–Parasite Co-evolutionary Processes

91

unpublished data). Reproductive isolation from the two Antarctic members of the C. osculatum complex (i.e. C. osculatum D and C. osculatum E) occurring sympatrically in the same definitive host (the Weddell seal), and from C. ogmorhini (s.s.) which occurs syntopically in the same host (the fur seal, Arctocephalus australis), was demonstrated by the lack of F1 hybrids and recombinant or introgressed individuals. The genetic relationships between C. mirounga and other congeneric taxa were later inferred from LSU rDNA sequences (Nadler et al., 2000) and from the mtDNA cox2 sequences analyses (Mattiucci et al., 2008b). No data on the larvae of this species which have been identified genetically are available.

4. PHYLOGENETIC SYSTEMATICS OF ANISAKID NEMATODES Allozyme data, typically consisting of allele frequencies obtained from starch gel electrophoresis of proteins, are an important source of characters for understanding the genetic structure of a population of anisakid nematodes and, consequently, for reconstructing phylogenies among conspecific populations and closely related species. Despite the increasing use of DNA sequence data in phylogenetics, allozyme data remain widely used in systematic and evolutionary studies of anisakid nematodes and have some advantages. For example, allozyme data consist of multiple unlinked nuclear loci, with each locus providing independent estimate of the species differentiation and phylogeny. Therefore, in contrast to results of nuclear and mitochondrial DNA data sets, allozyme data are less likely to be systematically missed by mismatches between gene trees and species trees. Furthermore, it is relatively cheap and easy to survey allozyme variation for a large number of individual anisakid nematodes. A problematic aspect using allozyme data, however, is that there is a longstanding and continuing controversy as to the preferred method for their phylogenetic analysis. Doubt remains as to the validity of allozyme data as a basis for inference of genetic relationship among related taxa. Although this issue is still debated in the literature, aspects could be addressed with empirical data recently obtained for anisakid nematodes. Genetic relationships inferred from allozyme markers have been largely obtained, in this group of parasites, by different methods such as phenograms from genetic distance methods, Neighbour Joining (NJ) from genetic matrix distances, spatial representation [Multidimensional Scale Ordination (MDS), Principal Component Analysis (PCA)] based on distance values or allele frequencies data, between several conspecific anisakid nematodes. From a phylogenetic point of view, anisakid species have been recognized based upon evidence of independent evolutionary lineages

92

Simonetta Mattiucci and Giuseppe Nascetti

reflected in the form of character states unique to the same individuals of a species. Indeed, genetically defined clades, corresponding to distinct evolutionary lineages, are consistent with their recognition as separate species. Congruence studies so far carried out on anisakid nematodes provided, in general, that clades supported by many different lines of evidence (allozymes, DNA sequences) can effectively be considered to be ‘known’ clades (as determined and supported by Felsenstein, 1985). As it has been presented in previous paragraphs, several molecular systematic studies of anisakid nematodes have employed more than one gene to define the taxonomic status of the species so far known. Indeed, generally, concordant evidence from independent gene analysis inspires much more confidence than the patterns that reflect separate lineage (and, thus, species existence) inferred from data obtained from a single gene. Congruence and incongruence evidences between allozyme data sets and not-allozyme data (i.e. DNA sequence analysis at both nuclear and mitochondrial genes) in anisakid nematode phylogenies are here reviewed. The opportunity to compare independent data sets supporting a given clade could affirm the existence of the clade with greater authority. In other words, shared hypothetical phylogenetic history seems to be the most likely explanation when congruence between diverse genetic data sets is observed. In this sense we are revising the phylogenetic relationships so far attempted for each group of anisakid nematodes belonging to the genera Anisakis, Pseudoterranova and Contracaecum. Indeed, phylogenetic analysis of partial 28S (LSU) sequences carried out by Nadler et al. (2005) of Anisakis, Pseudoterranova and Contracaecum taxa revealed strong support for the monophyly of the Anisakinae, Contracaecum plus Phocascaris, Pseudoterranova and Anisakis. Parsimony and ML analyses indicated that the Raphidascarididae, Contracaecum plus Phocascaris, and the Anisakinae (here considering only Pseudoterranova and Anisakis) are each monophyletic, the latter two groups with consistently strong bootstrap support at MP and ML analyses (Nadler et al., 2005).

4.1. Genetic relationships between Anisakis spp. Work performed on the phylogenetic studies within this genus was done initially using the unweighted pair group method using arithmetic averages (UPGMA phenograms) on the studied populations and species of Anisakis based on distance values inferred from DNei (Nei, 1972) and Dc (chord distance by Cavalli-Sforza and Edwards, 1967). Among the first extensive UPGMA phenograms were those generated by Mattiucci et al. (1997, 2001) and Paggi et al. (1998b) showing the genetic relationships between the species of Anisakis. This suggested that Anisakis is polyphyletic and highly heterogeneous, with a high genetic differentiation of A. physeteris and A. brevispiculata from the rest of the Anisakis taxa

Anisakid Nematodes and Host–Parasite Co-evolutionary Processes

93

recognized genetically. Other elaborations to infer genetic relationships between seven taxa of Anisakis (i.e. the A. simplex complex, A. typica, A. ziphidarum, A. brevispiculata and A. physeteris) included the MDS (Guttman, 1968) from chord-distance values (Dc, Cavalli-Sforza and Edwards, 1967), demonstrating that A. physeteris and A. brevispiculata are forming a well-distinct clade with respect to that formed by A. simplex complex, as well as from both A. typica and A. ziphidarum (see Mattiucci et al., 2002a). A consensus NJ tree was later inferred from Cavalli-Sforza and Edwards chord-distance values estimated at 19 enzyme loci between eight Anisakis taxa, including the new species A. paggiae (see Mattiucci et al., 2005). This showed consistently the existence of two distinct clades supported by high bootstrap values, including, respectively: (1) the three members of A. simplex complex, A. typica and A. ziphidarum within clade I and (2) A. physeteris, A. brevispiculata and A. paggiae within clade II (Fig. 2.2). Both A. ziphidarum and A. typica form two distinct subclades, closely related to the A. simplex species complex (Mattiucci et al., 2005). Maximum parsimony (MP) analysis of the ITS data set of 724 characters in seven taxa out of those genetically recognized in Anisakis, was performed by Nadler et al. (2005). This produced a strict consensus of the MP trees which also depicted A. physeteris plus A. brevispiculata as the sister group to the remaining Anisakis taxa. This clade was well supported in the MP bootstrap tree. According to this analysis, the remaining ingroup Anisakis were monophyletic in maximum likelihood (ML) and MP consensus trees, with 100% bootstrap support (Nadler et al., 2005). Later, Valentini et al. (2006) provided a phylogenetic hypothesis for all the nine taxa currently recognized species in Anisakis, using MP, NJ and Bayesian analysis (BA) inferred from mtDNA cox2 sequences data (629 bp). Phylogenetic trees generated show high congruence with the UPGMA and NJ trees obtained from allozyme data sets performed on the same species of Anisakis. An overall high congruence was indeed found between the tree topologies obtained from consensus NJ inferred from mtDNA cox2 sequence data and consensus NJ tree generated from allozyme data sets of the same taxa (Fig. 2.2). Both depicted the species A. physeteris, A. brevispiculata and A. paggiae as a sister group, highly supported, with respect to the other Anisakis taxa. In addition, A. paggiae appears to share a common ancestor with A. brevispiculata, and this is well supported by the NJ allozyme data. In addition, allozyme tree topology clearly demonstrated that Anisakis sp. formed a monophyletic group with A. ziphidarum; this was supported by the MP, BI and NJ of mtDNA cox2 sequences (Valentini et al., 2006) (Fig. 2.2). Finally, consistent tree topologies were observed between nuclear gene products and mitochondrial genes for the position occupied by A. typica as forming a separate clade—as depicted by mtDNA cox2-derived NJ (Valentini et al., 2006),

A (allozyme data)

B (mtDNA cox2) A. simplex (s.s.)

A. simplex (s.s.) 71

A. pegreffii

100 96

A. pegreffii

A. simplex C

99

100

A. simplex C

A. typica

A. typica

A. ziphidarum

73

A. ziphidarum

71

76 Anisakis sp.

Anisakis sp.

A. physeteris

A. paggiae 50

A. brevispiculata

A. brevispiculata A. physeteris

A. paggiae P. decipiens

70

(outgroup) P. decipiens Clade I Clade II

FIGURE 2.2 Genetic relationships among Anisakis spp. depicted by (A) neighbour-joining (NJ) tree inferred from chord-distance values (Dc, Cavalli-Sforza and Edwards, 1967) from allozyme data; (B) NJ inferred from K2P distance values obtained by mtDNA cox2 sequences analysis (data from Valentini et al., 2006). Bootstrap values 60 are shown at the internal nodes. Pseudoterranova ceticola as outgroup. The two data sets are congruent in depicting the existence of two main clades: (I) includes the A. simplex species complex [A. pegreffii, A. simplex (s.s.), A. simplex C], A. typica, A. ziphidarum and Anisakis sp.; (II) comprises the species A. physeteris, A. brevispiculata and A. paggiae. The species included in the two main clusters show a different larval morphology: type I (sensu Berland, 1961) in the first group and type II in the second one.

Anisakid Nematodes and Host–Parasite Co-evolutionary Processes

95

allozyme data, and inferred from ITS rDNA sequence analysis (Nadler et al., 2005) (Fig. 2.2). Therefore, phylogenies inferred from different sets of nuclear and mitochondrial data (i.e. allozymes, ITS rDNA sequences, mtDNA cox2) (Mattiucci et al., 2005; Nadler et al., 2005; Valentini et al., 2006) supported the group of species formed by A. physeteris, A. brevispiculata and A. paggiae as sister group to the remaining Anisakis taxa. Finally, the clade formed by the sibling species of the A. simplex complex received strong support in all the phylogenetic elaborations from allozymes (NJ), ITS-rDNA (MP) and mtDNA cox2 sequences (MP, NJ, BI). They were concordant to show a sister-group relationship for A. simplex (s.s.) and A. pegreffii, with A. simplex C as a sister to this clade (Mattiucci and Nascetti, 2006; Nadler et al., 2005; Valentini et al., 2006) (Fig. 2.2).

4.2. Genetic relationships between Pseudoterranova spp. Genetic relationships between Pseudoterranova species have included analyses based on allozyme genetic distance methods (Bullini et al., 1997, Paggi et al., 1991, 1998c). The allozyme analysis inferred from Dc chorddistance values between the five sibling species of the P. decipiens complex indicated the following topology: [P.decipiens (s.s.), P. azarasi, P. krabbei, P. bulbosa, P. decipiens E], with a close genetic relationship between P. decipiens (s.s.) and P. azarasi (Fig. 2.3). An UPGMA phenogram of uncorrected ITS rDNA distances (Zhu et al., 2002) was also produced for individual specimens, previously characterized by allozymes, belonging to the taxa: P. decipiens (s.s.), P. krabbei

P. decipiens (s.s.)

P. azarasi

P. bulbosa P. decipiens E 1.00

0.80

0.60

0.40

0.20

Nei’s D 0.00

FIGURE 2.3 Genetic relationships among Pseudoterranova spp. depicted by neighbourjoining (NJ) tree inferred from chord-distance values (Dc, Cavalli-Sforza and Edwards, 1967) from allozyme data (redrawn from Bullini et al., 1997).

96

Simonetta Mattiucci and Giuseppe Nascetti

P. krabbei, P. bulbosa, P. azarasi, P. cattani, and a taxon indicated as P. decipiens Ca1. This analysis showed the following relationship: [P. decipiens (s.s.), P. azarasi, P. bulbosa, P. cattani, P. krabbei, P. decipiens Ca1]. Zhu et al. (2002) also suggested that, because of the geographical origin, the P. decipiens Ca1 from Chaenocephalus aceratus, it is likely to correspond to P. decipiens E, but this correspondence has not been verified. An overall congruence was found between the topology produced by the allozyme phenogram (except for the absence of P. cattani from the allozyme analysis) produced by Bullini et al. (1997) and that obtained by both MP and ML analyses of ITS rDNA sequences by Nadler et al. (2005) performed on all the P. decipiens species complex. These topologies were congruent in depicting the close relationship between P. decipiens (s.s.) and P. azarasi. Nadler et al. (2005) showed that P. cattani and P. decipiens Ca1 are not closely related and they do not represent sister taxa with respect to the other taxa. According to those authors, this finding suggests a more complex evolutionary scenario than might be explained by a simple biogeographical scenario of their definitive hosts.

4.3. Genetic relationships between Contracaecum spp. A first phylogenetic hypothesis of the species included in the genus Contracaecum was that attempted using allozyme data on several species of Contracaecum from pinnipeds and fish-eating birds as definitive hosts, plus Phocascaris phocae and P. cystophorae (see Nascetti et al., 1990; Orecchia et al., 1986b). It was shown that Contracaecum species from seals were genetically most similar to each other rather than to the avian Contracaecum, with no alleles found in common between the two groups of species. In addition, it was clearly demonstrated that species of Phocascaris (i.e. P. phocae and P. cystophorae), despite their morphological characters (interlabia absent and/or reduced), were clustering within the clade formed by the species of the C. osculatum complex (interlabia present), which now includes five sibling species [C. osculatum A, C. osculatum B, C. osculatum (s.s.), C. osculatum D and C. osculatum E] (Nascetti et al., 1993; Orecchia et al., 1994) (see also Section 3.3.) Nadler et al. (2000) supported this first phylogenetic hypothesis using the nuclear-encoded large subunit ribosomal DNA sequences (LSUrDNA) for several taxa of Contracaecum (including seven species from pinnipeds and seven from fish-eating birds) and Phocascaris (two taxa), previously identified by allozyme markers. These data provided high support for the monophyly of all Contracaecum and Phocascaris of phocid seals. This finding is consistent with Berland’s hypothesis (1964) that such species form a group sharing the same life-cycle pathway. However, Berland’s proposal to recognize all species in phocid seals as Phocascaris, with all the species from birds as Contracaecum, was not congruent, according to

Anisakid Nematodes and Host–Parasite Co-evolutionary Processes

97

Nadler et al. (2000) with the inferred 28S rDNA trees produced, because it would result in a paraphyletic Contracaecum. Moreover, that phylogenetic elaboration, with respect to Berland’s hypothesis, would not depict avian parasites as monophyletic; some of them, for instance C. rudolphii species A and B, cluster in the same clade with the otariid parasite C. ogmorhini (see Nadler et al., 2000, 2005). A phenetic clustering (UPGMA), inferred from the allozyme data set (22 enzyme loci) only in Contracaecum spp. using pinnipeds as definitive hosts, showing their genetic relationship was produced by Nascetti et al. (1993). Later Arduino et al. (1995) and Bullini et al. (1997) produced a phenetic tree (UPGMA) showing the genetic relationships among six taxa, including the species of the C. osculatum complex [C. osculatum A, C. osculatum B, C. osculatum (s.s.), C. osculatum D, C. osculatum E] and C. radiatum. This depicted C. radiatum as the basal species of the first group of Contracaecum plus Phocascaris. Other graphical representations of the genetic relationships among the same group of Contracaecum taxa were those produced by MDS ordination obtained from Dc values in Orecchia et al. (1994) and the PCA analysis in Arduino et al. (1995), also including C. radiatum. A phenogram depicting the genetic differences observed in the entire ITS rDNA among specimens genetically identified previously by allozyme markers as belonging to the five taxa of the C. osculatum complex [i.e. C. osculatum A, C. osculatum B, C. osculatum baicalensis, C. osculatum (s.s.), C. osculatum D and C. osculatum E] was produced by Zhu et al. (2000b). This cluster analysis showed that C. osculatum A, C. osculatum D, C. osculatum E and C. osculatum baicalensis were genetically more similar to C. osculatum B than each was to C. osculatum (s.s.). Moreover, Zhu et al. (2000b) found C. osculatum (s.s.) the most genetically distinct taxon from the other members of the complex. This latter finding was concordant with that previously presented in Nascetti et al. (1993) and Orecchia et al. (1994). However, in the same genetic studies and phylogenetic representation, Zhu et al. (2000b) did not find any genetic differentiation in the ITS rDNA sequence analysis between C. osculatum D and C. osculatum E, previously detected by allozymes (Orecchia et al., 1994). Similarly, the lack of unequivocal nucleotide differences in any of three mtDNA sequences analysed (mtDNa cox1, ssrRNA and lsrrRNA) between C. osculatum D and C. osculatum E was also shown by Hu et al. (2001). These findings are inconsistent with allozyme data analysis (MAE) which, on the contrary, clearly demonstrated the reproductive isolation between C. osculatum D and C. osculatum in sympatry, with several allozyme loci fixed for alternative alleles in the two species. Moreover, the specific status of C. osculatum D and C. osculatum E was confirmed by the phylogenetic analysis, inferred from mtDNA cox2 sequences, which depicted the two taxa as two separate lineages (phylogenetic units) (Mattiucci et al., 2008b) (Fig. 2.4).

A (allozyme data)

B (mtDNA cox2)

98

C. osculatum A

C. osculatum B

C. osculatum E

C. osculatum (s.s.)

C. osculatum D

P. cystophorae C. osculatum A

C. o. baicalensis

C. osculatum D C. osculatum (s.s.)

90

96

62 94

C. o. baicalensis

C. osculatum B

90

95

96 99

C. osculatum E

85

P. cystophorae C. radiatum

88

C. radiatum

95 C. mirounga

C. mirounga

C. ogmorhini (s.s. )

C. ogmorhini (s.s. )

100 C. margolisi

C. margolisi P. ceticola

100

(outgroup)

P. ceticola

Clade I Clade II

FIGURE 2.4 Genetic relationships between species of Contracaecum and Phocascaris from pinnipeds depicted by (A) NJ tree inferred from chord-distance values (Dc, Cavalli-Sforza and Edwards, 1967) from allozyme data and (B) NJ inferred from mtDNA cox2 sequences analysis. Bootstrap values 60 are shown at the internal nodes. Pseudoterranova ceticola as outgroup. The two data sets are congruent in depicting the existence of two main clades: (I) includes the species of C. osculatum complex, Phocascaris cystophorae, C. radiatum, C. mirounga; (II) comprises C. ogmorhini (s.s.) and C. margolisi. The two phylogenetic trees show P. cystophorae nested in the subclade formed by the species of C. osculatum complex.

Anisakid Nematodes and Host–Parasite Co-evolutionary Processes

99

The overall concordant results with the allozyme data sets have been found in the tree topology inferred from the mtDNA cox2 sequences data (519 bp) ascertained for all the Contracaecum plus Phocascaris taxa so far recognized by allozyme markers [i.e. C. osculatum A, C. osculatum B, C. osculatum (s.s.), C. osculatum D, C. osculatum E, C. osculatum baicalensis, C. radiatum, C. mirounga, C. ogmorhini (s.s.), C. margolisi, Phocascaris cystophorae] (see Mattiucci et al., 2008b) (Fig. 2.4). A high congruence in the topology of the trees generated by a consensus NJ tree inferred from mtDNA cox2 sequences and that produced by a NJ based on Dc chord distance from allozyme markers was observed (Fig. 2.4). They are highly consistent in showing the existence of two well-supported clades: (1) one including the species of the C. osculatum complex, Phocascaris cystophorae, plus C. radiatum and C. mirounga (clade I), with the last two taxa being genetically closely related and forming a supported subclade; (2) the second clustering of the species C. ogmorhini (s.s.) and C. margolisi (clade II). Moreover, both elaborations from different data sets showed that the species P. cystophorae is nested in the subclade of the species of the C. osculatum complex. This is despite its morphological differences from the species of the C. osculatum complex. Allozyme NJ elaboration clearly depicted the species C. ogmorhini (s.s.) and C. margolisi as sister taxa with respect to the other Contracaecum taxa, and this was also shown by the mtDNA cox2 data. In addition, the Antarctic taxon C. osculatum D forms a monophyletic group with the other Antarctic member C. osculatum E, from which it was also found to be clearly genetically distinct (Table 2.10 and Fig. 2.4; see also Section 6). An overall congruence of this phylogenetic analysis was with that produced by Nadler et al. (2000) based on the LSU rDNA sequences analysis of seven Contracaecum species from pinnipeds [i.e. C. osculatum A, C. osculatum B, C. osculatum (s.s.), C. osculatum baicalensis, C. radiatum and C. mirounga]. Indeed, both NJ tree inferred from the full alignment data set using log-determinant distances and MP produced by Nadler et al. (2000) demonstrated the monophyly of Contracaecum plus Phocascaris from phocids. A strong support in all bootstrap trees also indicated for the subclades formed by C. radiatum and C. mirounga and that formed by C. osculatum A and C. osculatum baicalensis (see Nadler et al., 2005). These subclades were shown to be highly supported at MP and NJ analyses inferred from mtDNA cox2 sequences analysis (Mattiucci et al., 2006, 2008b) (Fig. 2.4). Furthermore, the trees inferred by all the methods, were congruent in demonstrating that Phocascaris spp. are clustering with the C. osculatum complex group of species and are well nested in the clade including the Contracaecum from phocids (see Mattiucci et al., 2008b; Nadler et al., 2000, 2005) (see also Section 6). Meanwhile, C. ogmorhini (s.s.) and C. margolisi cluster in a separate clade (Mattiucci et al., 2008b).

100

Simonetta Mattiucci and Giuseppe Nascetti

5. GENETIC DIFFERENTIATION IN ANISAKIDS Although the level of genetic differentiation is not a suitable parameter or a measurement for establishing species delimitation, values of genetic differentiation between taxa of anisakid nematodes inferred from nuclear loci analysis (19–22 enzyme loci) were found to be of a similar order. In sympatric or partially sympatric sibling species of anisakid nematodes, a significant deficiency or complete lack of some heterozygote classes at polymorphic loci strongly suggests that the sample being dealt with comprises distinct gene pools. This was widely demonstrated in samples of A. simplex (s.l.), P. decipiens (s.l.), and C. osculatum (s.l.) (see Mattiucci et al., 1997; Nascetti et al., 1986, 1993; Paggi et al., 1991). In allopatric anisakid nematodes, allozymes have provided reliable information on their specific status, with clear evidence of alternative and unique allozymes existing in allopatric populations. As the most common mode of speciation of endoparasites is the peripatric model (sensu Mayr, 1963, 1976), the build up of genetic divergence is generally a better predictor of whether two allopatric populations will interbreed upon recontact, than is conventional morphology. Indeed, according to the peripatric model of speciation, this involves geographical isolation of small populations whose genetic structure begins to differ from that of the parental population since the beginning of the process by genetic drift phenomena. In the Anisakis simplex complex and Contracaecum osculatum complex of species, genetic data strongly support the notion that adaptation to different hosts and speciation is strictly related to the geographical isolation of the hosts (see also Section 6.1). At least, in the case of anisakid nematodes, when Nei’s (1972) genetic distance between populations reaches values of
0.2–0.3, gene exchange is interrupted by intrinsic reproductive isolating mechanisms (RIMs). On the other hand, in anisakid nematodes, conspecific populations generally show similar allele frequencies, even when located thousands of kilometrers apart. Accordingly, their DNei values are quite low (0.0001– 0.002). Average genetic distance values inferred from allozymes within and among species of the genera are given in Tables 2.8–2.10.

5.1. Genetic differentiation at interspecific level Between sibling species of anisakid nematodes (sympatric and allopatric), the genetic differentiation, estimated at the nuclear level (allozymes), ranges on average, from values of Nei’s (1972) standard genetic distance (DNei) DNei  0.20 (as observed between the Arctic taxon C. osculatum A and the Antarctic species C. osculatum E) (Table 2.10) to DNei  0.90 (as found between the Antarctic taxon P. decipiens E vs the other species of

TABLE 2.8 Average of standard genetic distance by Nei (1972, DNei, below the diagonal) inferred from 20 enzyme-loci, and by Kimura2-parameter (Kimura, 1980, K2P, above the diagonal) inferred from 629bp of mtDNA cox2 between the species of Anisakis so far genetically detected

Species

A. simplex (s.s.)

A. simplex (s.s.) A. pegreffii A. simplex C A. typica A. ziphidarum Anisakis sp. A. physeteris A. brevispiculata A. paggiae

0.40 0.36 1.16 1.64 2.04 6.90 4.10 1

A. pegreffii

A. simplex C

A. typica

A. ziphidarum

Anisakis sp.

A. physeteris

A. brevispiculata

A. paggiae

0.05

0.06 0.06

0.14 0.14 0.14

0.12 0.12 0.13 0.13

0.13 0.13 0.14 0.12 0.09

0.14 0.14 0.14 0.15 0.13 0.15

0.17 0.17 0.17 0.18 0.14 0.16 0.11

0.14 0.14 0.15 0.15 0.12 0.13 0.12 0.13

0.37 1.45 1.99 2.63 8.30 6.11 1

1.14 1.62 1.92 7.40 5.54 1

1.67 1.62 4.77 3.49 1

0.68 1 1 1

1 1 1

0.95 1.06

(Data from: Mattiucci et al., (2001, 2002a, 2005), Mattiucci and Nascetti, 2006; Paggi et al., 1998b, and Valentini et al., 2006).

0.79

102

Simonetta Mattiucci and Giuseppe Nascetti

TABLE 2.9 Average values of genetic distance, DNei (1972), between species of the Pseudoterranova decipiens complex Species

P. krabbei

P. decipiens (s.s.) P. bulbosa P. azarasi P. decipiens E

0.51 0.93 0.56 0.91

P. decipiens (s.s.)

P. bulbosa

P. azarasi

0.60 0.38 0.96

0.64 0.97

0.93

Data from Bullini et al. (1997), Mattiucci et al. (1998, unpublished data), Paggi et al. (1991, 1998c).

P. decipiens complex), or DNei = 0.93 (as between P. krabbei and P. bulbosa) (Table 2.9). Similar high values were those observed in the comparison of Arctic members of the C. osculatum complex [i.e. C. osculatum D vs C. osculatum (s.s.), DNei = 0.70] and between members of the A. physeteris complex (i.e. A. physeteris vs A. paggiae and A. brevispiculata) (DNei  0.90) (see Mattiucci et al., 2005) (Table 2.8). The highest interspecific differentiation is that found at the allozyme level in the comparison of anisakid species that are morphologically distinct, such as the A. simplex complex and A. ziphidarum (see Paggi et al., 1998b) or the A. physeteris complex (Mattiucci et al., 2005). In the last group of species, spicule length of male specimens and the ventriculus are significantly shorter than those of the members of A. simplex complex. Moreover, in A. paggiae, the spicule length and ventriculus shape were demonstrated to be of diagnostic level not only from the members of A. simplex complex, but also with respect to genetically closely related species, that is, A. physeteris and A. brevispiculata (see Mattiucci et al., 2005). Analogously, a high genetic divergence at the allozyme level was also found between members of the C. osculatum complex with respect to C. mirounga (on average, DNei  1.05) (Mattiucci et al., 2008b) and to the C. ogmorhini species complex (on average, DNei  1.30) (Mattiucci et al., 2003, 2008b), which are morphologically distinct. The highest level of differentiation at allozyme level was found between the A. simplex species complex with respect to the A. physeteris species complex (Mattiucci et al., 2005) (Table 2.8). Nuclear and mitochondrial DNA polymorphisms among species of anisakid nematodes demonstrate significantly high genetic variation among sibling species and morphospecies of those anisakid nematodes previously characterized genetically by allozyme markers. Indeed, it has recently been shown that most of the genetic diversity is strongly structured between species rather than within species. Pairwise comparisons were estimated of the level of sequence differences at the ITS rDNA between 14 Anisakis samples by Nadler et al. (2005). Sequence differences

TABLE 2.10 Average of standard genetic distance by Nei (1972, DNei, below the diagonal) inferred from 20 enzyme-loci, and by Kimura2-parameter (Kimura, 1980, K2P, above the diagonal) inferred from 519 bp of mtDNA cox-2 between Contracaecum spp. from pinnipeds. Species

COSA

C. osculatum A (COSA) C. osculatum B (COSB) C. osculatum (s.s.) (COSS) C. osculatum D (COSD) C. osculatum E (COSE) C. baicalensis (CBAI) P. cystophorae (PCYS) C. radiatum (CRAD) C. mirounga (CMIR) C. ogmorhini (s.s.) (COGM) C. margolisi (CMAR)

0.41 0.57 0.23 0.20 0.30 0.45 0.72 1.00 1.30 1.35

COSB

COSS

COSD

COSE

CBAI

PCYS

CRAD

CMIR

COGM

CMAR

0.09

0.10 0.09

0.09 0.09 0.11

0.09 0.10 0.10 0.05

0.06 0.08 0.10 0.09 0.09

0.09 0.10 0.07 0.11 0.11 0.09

0.14 0.14 0.16 0.13 0.15 0.16 0.15

0.13 0.13 0.15 0.13 0.13 0.12 0.15 0.09

0.16 0.15 0.14 0.16 0.14 0.15 0.16 0.14 0.14

0.15 0.15 0.15 0.16 0.15 0.16 0.16 0.15 0.15 0.05

0.80 0.38 0.36 0.27 0.83 0.76 1.06 1.35 1.40

0.64 0.53 0.41 0.46 1.30 1.13 1.37 1.42

0.25 0.60 0.68 0.78 1.15 1.40 1.39

0.50 0.51 0.67 0.86 1.48 1.45

(Data from: Nascetti et al., 1993; Orecchia et al., 1994; Mattiucci et al., 2008b and unpublished).

0.64 0.79 1.20 1.70 1.70

0.96 1.07 1.90 1.89

0.95 1.60 1.75

1.81 1.80

0.30

104

Simonetta Mattiucci and Giuseppe Nascetti

(absolute differences) of 0–7 existed among individuals belonging to the same species, whereas it ranged from 37 to 149 between individuals belonging to distinct species (Nadler et al., 2005). At the mtDNA level, the genetic divergence estimation among the nine Anisakis taxa, previously characterized on the basis of allozyme data, was calculated at Kimura-2-parameters distance values (K2P) inferred from sequence analysis (629 bp) of the mtDNA cox2 gene by Valentini et al. (2006) (Table 2.8). The mtDNA cox2 fragment in all the Anisakis was found to be A+T rich (60.7%, 64.9% and 74.7% at the first, second and third positions, respectively). The lowest level of interspecific genetic distance was found between sibling species of the Anisakis simplex complex (on average, K2P  0.05). Whereas, values of K2P = 0.14 and 0.13 were observed when comparing the A. simplex complex, on average, with A. typica and A. ziphidarum. Similar values (on average K2P  0.13) were obtained when A. physeteris, A. brevispiculata and A. paggiae are compared to A. ziphidarum (see Valentini et al., 2006) (Table 2.8). At the amino acid level, a total of 31 variable positions were identified in the mtDNA cox2 gene for Anisakis spp. The average variation ranged from 2.4% between the sibling species of A. simplex complex to as high as 7.7% between morphologically distinct species (i.e. A. physeteris, A. brevispiculata and A. paggiae vs the A. simplex complex (Valentini et al., 2006)). Single strand conformation polymorphism (SSCP) analysis of the ITS rDNA, performed by Zhu et al. (2002), has revealed that, while no variation in single stranded ITS profiles existed within each of the five sibling species of P. decipiens complex (with the exception of a slight microheterogeneity evidenced in P. bulbosa and P. cattani), SSCP analysis of the ITS-2 amplicons allowed significant differentiation between them (Zhu et al., 2002). Pairwise comparison of the ITS sequences revealed nucleotide differences ranging from 0 to 6.8%, which were within the range observed in other complex members (0–2.3% in the C. osculatum complex) (see Zhu et al., 2000b). Sequence variation at the same internal transcribed spacers of ribosomal DNA (ITS-rDNA) within and among members of the Contracaecum osculatum complex and SSCP was used to screen the ITS-1 and ITS-2 amplicons separately for sequence variation within and among individuals (Zhu et al., 2000b). The G+C contents of the sequences obtained were 45.7% (ITS-1) and 42.0–43.5% (ITS-2). However, no variation in single-strand profiles was detected between/among samples representing a taxon by Zhu et al. (2000b), except for slight nucleotide polymorphism detected in the ITS for C. osculatum A and C. osculatum B. Indeed, there was no variation in the length of the ITS-1 (499 bp) or ITS-2 (262 bp) sequence for any of the individuals examined, and no-intraspecific variation in the sequence was observed. Whereas, significant inter-taxon

Anisakid Nematodes and Host–Parasite Co-evolutionary Processes

105

differences in SSCP profiles were detected between C. osculatum A, C. osculatum B, C. osculatum C and C. osculatum baicalensis. This revealed a reliable genetic differentiation of these taxa from one another, except in the case of C. osculatum D versus C. osculatum E, which exhibited identical sequences (Zhu et al., 2000b). Each of the other four taxa had distinct sequences with interspecific differences ranging from 0.3 to 2.3% (Zhu et al., 2000b). C. osculatum (s.s.) was the most distinct taxon with respect to all the other members. Comparison among the taxa revealed 22 variable nucleotide positions in the ITS sequence. Each of the taxa could be identified by the presence of 1–10 fixed nucleotide differences, except in the case of C. osculatum D and C. osculatum E, which were found at the ITS rDNA gene to show the same nucleotide sequence. At the mitochondrial level, the genetic divergence estimated from sequence analysis (519 bp) of the cox2 gene was given for the five members of the C. osculatum complex [i.e. C. osculatum A, C. osculatum B, C. osculatum (s.s.), C. osculatum D and C. osculatum E] with respect to C. osculatum baicalensis, C. radiatum, C. mirounga, Phocascaris cystophorae and to the otariid parasites, C. ogmorhini (s.s.) and C. margolisi (see Mattiucci et al., 2008b). Genetic divergence has been shown to range from K2P = 0.05 (C. osculatum D vs C. osculatum E) to K2P = 0.11 [C. osculatum (s.s.) vs C. osculatum D] between sibling species of the C. osculatum complex. Higher values of K2P were those estimated between morphologically differentiated species such as in the comparison of C. radiatum and the species of C. osculatum complex (on average K2P  0.14), or C. osculatum complex versus C. ogmorhini complex species (on average K2P  0.15) (Mattiucci et al., 2008b) (Table 2.10). However, despite morphologically distinct characters between Phocascaris spp. and the species so far included in the C. osculatum complex, P. cystophorae is genetically very closely related to this group of species (on average K2P  0.10) (Mattiucci et al., 2008b) (Table 2.10) (see also Section 4.3). The Contracaecum mtDNA cox2 gene was found to be A+T rich. This, however, is consistent with that found in other anisakid nematodes, such as Contracaecum parasites of fish-eating birds (Mattiucci et al., 2008c). Similar distance values were those estimated at the same mitochondrial gene (mtDNA cox2) among species of Contracaecum parasitic in fish-eating birds. Indeed K2P ranged from 0.08 between the two sibling species, C. rudolphii A versus C. rudolphii B, to 0.13 between the two sibling species and species morphologically well differentiated (i.e. the C. multipapillatum complex) (Mattiucci et al., 2008c, unpublished data). The genetic divergence among the anisakid taxa so far evaluated at the mtDNA cox2 locus is of the same order as found between other nematode species (Anderson et al., 1998; Blouin et al., 1998; La Rosa et al., 2001; Pozio and Murrel, 2006; Zarlenga et al., 1998).

106

Simonetta Mattiucci and Giuseppe Nascetti

5.2. Genetic differentiation at the intraspecific level and gene flow In Anisakis spp., most of the genetic diversity has been found within populations rather than among populations (see also Section 8.2). Indeed, for instance, at the intraspecific level, the values of DNei found between populations of A. simplex (s.s.) ranged from 0.003 to 0.0038. Such a pattern of genetic distance is related to that of geographical distance (isolation by distance). Average DNei values is 0.007 among Atlantic populations, DNei  0.015 among Pacific ones, and DNei  0.025 between the two groups. The highest value (DNei  0.038) was found between the most distant populations, from the Baltic Sea and off Hokkaido Island, Japan. This finding is mainly a consequence of the differences in allele frequencies at the loci Sod-2, fEst-2, Lap-2 and Pep B in populations of A. simplex (s.s.), compared with those from the other localities (Mattiucci et al., 1997, 1998). At the interpopulation level, A. pegreffii showed the lowest values of Gst (average 0.02, Mattiucci et al., 1997) and DNei is, on average  0.005. This finding indicates remarkable genetic homogeneity of A. pegreffii throughout its wide range, from the Austral region to the Mediterranean Sea. A more consistent intraspecific distance (DNei, on average, 0.045; Gst = 0.10) was found in A. simplex C between the population from the Canadian Pacific and those from Austral region (South Africa, New Zealand, Tasman Sea and sub-Antarctic area) (Table 2.8). This is mainly revealed by the locus Sod-1, which markedly differentiated the sample from the Canadian Pacific and those from the Austral region (Mattiucci et al., 1997). Also in the case of A. typica (s.l.), remarkable genetic homogeneity was observed in larval and adult samples despite being geographically quite distant. Interpopulational genetic distances values detected among populations from Somali, Mediterranean and Brazilian waters were low, ranging from DNei  0004 (eastern Mediterranean vs Somali waters) to DNei  0.010 (Brazilian vs Somali). The average value of Fst among all the populations was 0.04 (Mattiucci et al., 2002a). Similar low levels of population structuring were those found among Boreal and Austral populations of other taxa, such as DNei  0.002 within A. ziphidarum (see Paggi et al., 1998b), DNei  0.008 and DNei  0.009, respectively, in A. brevispiculata and A. physeteris (Mattiucci et al., 2001) and, finally DNei  0.005 in A. paggiae, despite the geographical distance between the samples, indicating high levels of gene flow in all of these Anisakis taxa (Table 2.8). Similar values, at the intraspecific level, were found in the comparison of populations belonging to the sibling species of the P. decipiens complex. The four members of this complex from the Arctic Boreal region showed DNei among populations ranging from, on average 0.001 (in P. krabbei) to DNei  0.005 [in P. decipiens (s.s.)]. Indeed, the lowest genetic heterogeneity was found within P. bulbosa, despite the geographical isolation of the

Anisakid Nematodes and Host–Parasite Co-evolutionary Processes

107

populations (NE and NW Atlantic, Pacific) with an average of 0.004. The highest intraspecific differentiation was detected within P. azarasi, for example, the sample from Anadyr (Bering Sea) showed a value of DNei = 0.026 with respect to conspecific Japanese samples (Mattiucci et al., 1998; Paggi et al., 1998c). Both P. decipiens (s.s.) and P. azarasi showed a strict correlation between geographical and genetic distance, whereas higher levels of gene exchange were detected within P. bulbosa. Analogous average values and ranges of genetic diversity between conspecific populations were reported for the members of the C. osculatum complex, from both Arctic and Antartic regions. At the intraspecific level, the values of DNei within C. osculatum A ranged from 0.002 to 0.017, with an average of DNei  0.009. Average DNei values were 0.004 within Northern Atlantic populations, while they were 0.008 among the conspecific populations and 0.018 between the two groups (Mattiucci et al., 1998; Paggi et al., 1998c). This demonstrates a geographical pattern in the genetic variation, although less marked than that observed in A. simplex (s.s.). Among conspecific populations of C. osculatum B, a lower level of differentiation was found DNei  0.005 on average, with a range of values from 0.001 to 0.014, the last value being reported among Pacific and Atlantic populations (Paggi et al., 1998c). A lower degree of interpopulational genetic differentiation was found between populations of the two Antarctic taxa, C. osculatum D and C. osculatum E, detected in definitive and fish hosts from the Weddell and Ross Seas, on average, DNei  0.003 and DNei  0.002, respectively. The highest levels of interpopulational distance between anisakid populations were those found within the otariid parasite C. ogmorhini (s.s.), ranging from DNei = 0.021 to DNei = 0.029 between samples from Argentine, South African and New Zealand waters (on average, DNei = 0.024) (Mattiucci et al., 2003). Similar values are generally found among races in different organisms (Ayala, 1975; Bullini and Sbordoni, 1980). Allozyme data sets have indicated that genetic variation of anisakids is generally not structured in geographical races or subspecies. Such low values of Nei’s D between populations located thousands of kilometres apart are likely to be caused by the homogenizing effects of gene flow, enhanced by the high dispersal capacity of intermediate/paratenic and definitive hosts. Indeed, levels of interpopulational gene flow, indirectly estimated from allele frequencies and Fst values, have been reported in these marine ascaridoid nematodes to be at high levels (Table 2.11). Despite the geographical distance between sampling locations, high rates of gene flow were detected within species belonging to three genera. The highest values were those observed in the species included in the Anisakis simplex complex. As we have illustrated above (Section 3.1), these worms mature in several species of ‘oceanic dolphins’ whose large geographical distribution could maintain the high level of gene flow observed in these anisakid nematodes (Table 2.11). Similarly, the high

108

Simonetta Mattiucci and Giuseppe Nascetti

TABLE 2.11 Values of intraspecific gene flow, estimated in anisakids from the standardized variance of allele frequencies, Fst (Crow and Aoki, 1984; Slatkin and Barton, 1989); m is the fraction of immigrant individuals in a population of effective size N

Species

A. simplex (s.s.) A. pegreffii A. simplex C A. typica A. physeteris A. brevispiculata P. krabbei P. decipiens (s.s.) P. bulbosa C. osculatum A C. osculatum B C. osculatum (s.s.) C. osculatum D C. osculatum E C. ogmorhini (s.s.)

Gene flow (Nm)

8.00 15.00 9.00 6.00 5.10 6.20 4.38 3.66 9.36 5.20 4.90 3.90 4.60 6.10 4.10

References

Mattiucci et al., 1997 Mattiucci et al., 1997 Mattiucci et al., 1997 Mattiucci et al., 2002 Unpublished data Unpublished data Mattiucci et al., 1998; Paggi et al., 1991 Mattiucci et al., 1998; Paggi et al., 1991 Mattiucci et al., 1998; Paggi et al., 1991 Mattiucci et al., 1998; Nascetti et al., 1993 Nascetti et al., 1993 Nascetti et al., 1993 Orecchia et al., 1994, unpublished data Orecchia et al., 1994, unpublished data Mattiucci et al., 2003

mobility and dispersal capacity of their intermediate/paratenic hosts (fish) could enhance the high degree of gene exchange. Such figures are similar to those found for the Arctic-Boreal members belonging to the C. osculatum and P. decipiens complexes maturing in pinnipeds (Table 2.11). These findings confirm that the complex life cycle of these nematodes does not limit gene exchange, but enhances it through the high mobility of the different hosts, such as fish and marine mammals (see also Section 7). On the other hand, the possible correlation between gene exchange and host mobility has also been suggested for other nematodes with different life cycles, as inferred by other genetic markers (Anderson et al., 1998; Blouin et al., 1995, 1998; Criscione et al., 2005).

6. HOST–PARASITE COPHYLOGENY Uncovering the phylogenetic history of anisakid parasites and their hosts represents a steadily advancing part of their systematics and ecology. Thus, cophylogeny mapping has become a recent development in the study of the relationships among and between ecologically related anisakid nematodes with respect to their hosts. Generally, in cophylogeny studies, a dependent phylogeny, such as that of a taxonomic group of

Anisakid Nematodes and Host–Parasite Co-evolutionary Processes

109

parasites, is mapped onto an independent one, such as that of their hosts. Co-evolutionary studies mostly rely on morphological or molecular data which are now becoming increasingly available in both parasites and their hosts, allowing detailed comparison of parasite phylogenies with those proposed for their hosts. Results of such studies have demonstrated how the phylogeny of the parasites comes to ‘mirror’ that of their hosts (sensu Eichler, 1948; Fahrenholz, 1913). In cophylogeny mapping, the map is constructed to provide the best possible explanation for the phylogenies and thus determine whether the parasites have co-diverged with their hosts, or whether they have undergone host switching, duplication by independent speciation or even whether parasites had extinction events (Charleston, 2003; Desdevises, 2007). Another method of uncovering ancient associations between dependent and independent taxonomic groups is the ‘biogeographical method’ also known as Brooks’ Parsimony Analysis (Brooks, 1990; Brooks and Hoberg, 1981). In this model, hosts are considered as ‘geographical regions’ and their parasites are their ‘endemic species’, undergoing codivergence in vicariant speciation, loss by extinction, ‘missing the boat’ (sensu Paterson et al., 1993), sympatric speciation and horizontal transfer by migration. Similar investigations have been made into host–parasite associations between anisakid nematodes and their definitive hosts. Phylogenetic trees, mostly inferred from molecular genetic data sets, are now available for these parasites and for some taxonomic groups of their definitive hosts. Comparison of these trees has allowed identification of cophylogenetic events. Mattiucci and Nascetti (2006) have recently provided evidence for such cophylogenetic events in the host–parasite systems of Anisakis spp. and cetaceans (their main definitive hosts). The presence of the two main clades, as presented in the phylogenetic relationships between Anisakis spp. (Section 4.1), is supported also by ecological data and specific host– parasite relationships. Phylogenetic relationships proposed elsewhere and here revised for species of Anisakis ‘mirror’, in several host–parasite associations, that of their definitive hosts (Arnason et al., 2004; Cassens et al., 2000; Milinkovitch, 1995; Nikaido et al., 2001). The sperm whales, Physeter catodon, Kogia breviceps and K. sima are the main definitive hosts for A. physeteris, A. brevispiculata and A. paggiae, respectively (Mattiucci and Nascetti, 2006; Mattiucci et al., 2001, 2005) (Figs. 2.2 and 2.5). Several oceanic dolphins in the Delphinidae, Arctic dolphins in the Monodontidae and porpoises in the Phocoenidae (Table 2.4) are hosts of the species of the A. simplex complex and of A. typica (see Mattiucci et al., 1997, 2002a, 2005). The beaked whales Ziphius cavirostris, Mesoplodon layardii, M. mirus and M. grayi are hosts of A. ziphidarum (see Paggi et al., 1998b) and Anisakis sp. (see Valentini et al., 2006) that are partitioned in clade I in the parasite phylogenetic tree (Fig. 2.3). The phylogeny of cetaceans

Host Bottlenose dolphin

Parasite A. pegreffii 71

F

Short-finned pilot whale

Mage19

E

A. simplex (s.s.)

Isi14 Isi36 Isi36 Mago24 Mago26 Mago32

G

Narwhal

Amz13 Bando1

D

Amazon river dolphin H

60

A. simplex C

Amz11

La plata dolphin

Mago8 Mago13

Yangtze river dolphin

I

C

Tuti24 Tuti35

Mago21 Mago22

A. typica 73

Baird’s beaked whale Mesoplodon sp.

B

A. ziphidarum

76

Anisakis sp.

Sperm8 Sperm28 Sperm47

Ganges river dolphin J

A Bando1 SP316

100

Dall’s porpoise

K Sp9

Sp2

Pygmy sperm whale

Pm72 Pm52 M11

A. paggiae A. brevispiculata

Sperm whale L Hump20 Hump203 aaa792

70

A. physeteris Humpback whale Fin whale

(outgroup) P. ceticola Dolphins

Minke whale

Hosts of Anisakis spp.

Ziphiids Physeterids

FIGURE 2.5 Tanglegram of phylogenies of Anisakis spp. and their cetacean hosts. It shows the phylogeny of a group of extant cetaceans (adapted from Nikaido et al., 2001) mapped into the phylogeny of Anisakis spp. inferred from mtDNA cox2 sequence analysis. Lines depict the observed host–parasite co-speciation events; the dotted line indicates possible host-switching events (redrawn from Mattiucci and Nascetti, 2006).

Anisakid Nematodes and Host–Parasite Co-evolutionary Processes

111

representing all extant cetacean families has recently been proposed by Milinkovitch (1995) based on mtDNA genes (12S, 16S and cytb partial sequences) and on myoglobin sequences; by Cassens et al. (2000) inferred from sequence analyses of 12S, 16S and cytb; by Nikaido et al. (2001) based on retroposon analysis and by Arnason et al. (2004) based on analyses of complete mitochondrial genomes. According to the phylogenetic hypothesis proposed by Arnason et al. (2004), the Cetacea group splits into monophyletic Mysticeti (baleen whales) and monophyletic Odontoceti (toothed whales). The Odontoceti diverged into the four extant lineages, Physeteridae (sperm whales: represented by the sperm and pygmy sperm whales), Ziphiidae (beaked whales), Platanistidae (Indian river dolphins) and Delphinoidea (encompassing the families Iniidae, Monodontidae, Phocoenidae and Delphinidae). Phylogenetic trees provided by Nikaido et al. (2001) and Arnason et al. (2004) were congruent in depicting the branching order of the extant cetacean lineages, where the sperm whale and the pygmy sperm whales (belonging to the families Physeteridae and Kogiidae) represent basal taxa, followed by the beaked whales (belonging to the family Ziphidae) and the freshwater and marine dolphins as the most derived (Fig. 2.5). In accordance with those analyses, the branching order so far proposed for the Anisakis taxa showed that nematodes from the sperm whale and pygmy sperm whales (i.e. A. physeteris, A. brevispiculata and A. paggiae) always occupy a basal lineage, always well supported, followed by those parasitizing the beaked whales (A. ziphidarum and Anisakis sp.). The species of the A. simplex complex and A. typica, parasites of delphinoids, are the most derived (Mattiucci and Nascetti, 2006) (Fig. 2.5). Elaboration of these empirical results to assess the global congruence between host and parasite trees gained by ParaFit (Legendre et al., 2002) was statistically significant (P < 0.05). Individual host–parasite associations which contributed more to the cophylogenetic cetacean—Anisakis spp. mapping were represented by those between A. physeteris–Physeter catodon, A. brevispiculata–Kogia breviceps and A. ziphidarum–Mesoplodon spp., suggesting host–parasite co-speciation events. Whereas, a lower significant contribute to the total test was that formed by the host–parasite association A. simplex (s.s.)–Balenoptera acutorostrata, suggesting a possible host-switching event (our unpublished data). Such parallelism between host and parasite phylogenies is also attempted for the group of Contracaecum taxa and their definitive hosts, the pinnipeds of the Families Phocidae and Otariidae. The presence of the two main clades, as presented in the phylogenetic relationships among Contracaecum spp. (Fig. 2.4), is also supported by the ecological data concerning host preference (see also Section 7.1) and specific host–parasite relationships. Phylogenetic relationships so far proposed and here reviewed for species of Contracaecum parallel that recently reported for their definitive hosts based on molecular data (Arnason et al., 1995;

112

Simonetta Mattiucci and Giuseppe Nascetti

Deme´re´ et al., 2003). Several phocid seals (true seals) in the Phocinae are hosts of the species of the C. osculatum complex and Phocascaris spp. in the Arctic Boreal region. Moreover, seals in the Monachinae are hosts for C. mirounga and C. radiatum in the Antarctic and sub-Antarctic region. These anisakids are included in clade I in the parasite phylogenetic tree (Mattiucci et al., 2008b) (Figs. 2.4 and 2.6). Whereas, the otariid species in the Otariinae (sea lions) Zalophus californianus, and in the Arctocephalinae (fur seals) Arctocephalus spp. are hosts of the C. ogmorhini species complex [C. margolisi and C. ogmorhini (s.s.)]. These anisakids are included in clade II in the parasite phylogenetic tree (Mattiucci et al., 2008b) (Figs. 2.4 and 2.6). Although a complete species-level phylogeny for pinnipeds including fossil and extant taxa is as yet, unavailable, a molecular assessment of pinniped relationships was performed by Arnason et al. (1995) using the complete sequences of the mitochondrial cytochrome b gene (mtDNA cytb) of the Phocidae, Odobenidae and Otariidae. Later, Deme´re´ et al. (2003) used a composite tree inferred from the basic topology of generic level, morphological and molecular data, fossil taxa and consensus phylogeny of the phocid subfamilies to propose an integrated hypothesis for pinniped evolutionary biogeography. According to that data elaboration, the Pinnipedia includes three major monophyletic clades: (1) the Otariidae (fur seals and sea lions), (2) the Odobenidae, and (3) the Phocidae (true seals), plus the extinct desmatophocids. In this combined tree, the fur seals and sea lions comprising the Otariinae (Zalophus californianus) and the Arctocephalinae (Arctocephalus spp. from the southern hemisphere), are represented as well-supported basal groups (Arnason et al., 1995; Deme´re´ et al, 2003). In accordance with that analysis, the branching order so far proposed for the Contracaecum taxa showed that nematodes from the Otariidae [i.e. C. ogmorhini (s.s.) from Arctocephalus spp. and C. margolisi from Zalophus californianus] always occupy a basal lineage of the parasite phylogenetic tree, with the species of the C. osculatum complex from the Phocinae (true seals) as the most derived (Fig. 2.6). Speciation of the members of C. osculatum complex is apparently related to their geographical isolation, through that of their hosts, as well as to a rapid host–parasite adaptation and co-evolution. Such processes apparently occurred in different times during the Plio-Pleistocene, when extreme climatic variation took place. The genetic relationships found between the members of the C. osculatum complex suggest that the evolutionary divergence of the most differentiated species [C. osculatum (s.s.)] started more than 3 million years ago, in a Pleistocene refuge (the Baltic Sea). As to the other C. osculatum species, their evolutionary divergence probably took place during the Pleistocene, when the complex achieved a distribution over both polar regions. This process involved two distinct colonizations of the marine Antarctic region by ancestors of the northern hemisphere, giving rise to C. osculatum D and C. osculatum E,

Parasite

Host P. vitulina vitulina P. vitulina richardsi

97 65

P. largha H. grypus

C. osculatum B C. osculatum (s.s.)

95

Ph. cystophorae C. osculatum A

P. hispida C. o. baicalensis

90 87

62 94

89

P. fasciata

C. osculatum D

96 99

P. groenlandicus E. barbatus

74

80

L. weddellii

85

C. osculatum E C. radiatum

95

M. schauinslandi H. leptonyx

97

M. leonina

C. mirounga

C. ursinus Z. californianus

100 100

88

91

A. gazella

C. margolisi

100

C. ogmorhini (s.s.)

A. forsteri (outgroup) P. ceticola

97

93 100

O. rosmarus Brown bear Polar bear

American black bear Domestic cat Lion

Hosts of Contracaecum spp.

Phocinae Monachinae Otariidae

Tiger

FIGURE 2.6 Tanglegram of phylogenies of Contracaecum spp. and their pinnipeds hosts. It shows the phylogeny of a group of extant pinnipeds (adapted from Arnason et al., 1995) mapped into the phylogeny of Contracaecum spp. inferred from mtDNA cox2 sequence analysis (adapted from Mattiucci et al., 2008b). Lines depict possible host–parasite co-speciation events; the dotted line indicates possible hostswitching events.

114

Simonetta Mattiucci and Giuseppe Nascetti

both parasites of Leptonychotes weddellii (see Bullini et al., 1997). This hypothesis seems to fit with the evolutionary biogeography of a pinnipedimorph hypothesis based on both dispersal and vicariant events in the context of a species-level phylogenetic framework proposed by Deme´re´ et al. (2003). This hypothesis supports an eastern North Pacific origin during the late Oligocene coincident with start of glaciation in Antarctica. During the late Miocene, pinnipedimorphs remained restricted to the eastern North Pacific, where they began to diversify. Fur seals remained restricted to the North Pacific until the late Pliocene, with a dispersal and rapid speciation in the Southern Ocean during the Pleistocene. The phocine seal diversification took place in the Arctic and North Atlantic during the late Miocene with a subsequent dispersal into the Paratethys and Pacific during the Pleistocene. Finally, the monachine seals, including Mirounga leonina and Leptonychotes weddellii, seem to have the southern hemisphere as the centre of diversification (Deme´re´ et al., 2003). The mode of speciation which apparently fits well with the anisakid nematodes is the peripatric model proposed by Mayr (1963, 1976) (see also Section 5). This involves the geographical isolation of small populations whose genetic structure begins to differ from the parental one by different genetic mechanisms. In the case of the Anisakis spp., and the C. osculatum and P. decipiens species complexes, molecular genetic data strongly suggest that adaptation to different hosts and speciation is related to the geographical isolation of the hosts. Such processes apparently occurred in different times from the lower Miocene to Pliocene, and Pleistocene, when extreme climatic variation took place. During glacial maxima (a period also of lowest sea level), smaller populations of hosts and their endoparasites would have remained isolated in marine refuges, promoting genetic divergence and co-adaptation. Then during interglacial periods, geographical ranges might have expanded, favouring host range expansion (Bullini et al., 1997). Similar co-evolutionary processes have been proposed by Hoberg (1992, 2005) for other host–parasite interactions, involving Holarctic cestodes and their definitive hosts (fish-eating birds and marine mammals).

7. HOST PREFERENCE, ECOLOGICAL NICHE AND COMPETITION Molecular genetic markers of the so far genetically characterized species of anisakid nematodes provide a rapid, precise means to screen and identify several marine mammals, fish and squids that serve as their definitive and intermediate and or/paratenic hosts. These data give important information regarding the life-history traits and epizootiological aspects of these parasites.

Anisakid Nematodes and Host–Parasite Co-evolutionary Processes

115

Marked differences in definitive and intermediate host preferences were detected among the species belonging to Anisakis, members of Pseudoterranova decipiens complex and species of Contracaecum from pinnipeds. They have shown distinct host preferences, with respect to their definitive and intermediate hosts, often exhibiting differential distribution in various host species and/or with different ecology. Patterns of differential distribution are the product of ecological processes and therefore interactions between parasite populations could promote and even enhance the complexity of multi-species assemblages observed in a given host (definitive and intermediate). Such differences in the life history and host preference between related species of anisakid nematodes appear to be due to differential host–parasite co-adaptation and co-evolution as well as to interspecific competition. This acts either in reducing the range of potential hosts or promoting a differential use of resources from a single individual host in sympatric conditions. In other words, the realized niche in some anisakid species could be more restricted than the fundamental niche. Some examples of such ecological processes have been found in anisakids of the P. decipiens complex and C. osculatum complex and are described below. In the North Atlantic and Pacific Oceans (Fig. 2.7), P. bulbosa is the only species present in the bearded seal, Erignathus barbatus, both in the eastern and western parts of the Atlantic Ocean, and also in the Pacific (Table 2.5), whereas in the North Eastern Atlantic waters P. decipiens (s.s.) and P. krabbei occur sympatrically as adults and often also syntopically in the same individual host in the common seal, Phoca vitulina, and in the grey seal, Halichoerus grypus (see Mattiucci et al., 1998; Paggi et al., 1991). However, statistically significant differences in relative numbers have been observed between P. decipiens (s.s.) and P. krabbei identified by genetic markers (allozymes) in those hosts, where P. decipiens (s.s.) predominates about tenfold in the common seal, while P. krabbei is the main prevalent species in the grey seal. On the other hand, in the western Atlantic, where P. krabbei is absent, P. decipiens (s.s.) occurs in equal proportions in both grey and common seals (Nascetti, 1992; Paggi and Bullini, 1994) (Fig. 2.7). This finding was also observed by Brattey and Stenson (1993), using the same genetic markers (diagnostic allozymes), to identify to species level anisakid nematodes in seal hosts from Newfoundland and Labrador. The bearded seal was found to be parasitized by P. bulbosa (under its former name, P. decipiens C, by Brattey and Stenson, 1993), while both the grey and common seal were found heavily infected by P. decipiens (s.s.) (under its former name, P. decipiens B, by Brattey and Stenson, 1993). This finding could reflect differential adaptation of the two parasites, which have evolved allopatrically. After their secondary contact and, consequently, their sympatric occurrence in the North East Atlantic seal hosts, interspecific competition between the two

North Atlantic Ocean

Halichoerus grypus

Erignathus barbatus

Pagophilus groenlandicus

Pseudoterranova bulbosa

Contracaecum osculatum A

Pseudoterranova decipiens (s.s.)

Contracaecum osculatum B

Pseudoterranova krabbei

Contracaecum osculatum (s.s.)

Phoca vitulina

FIGURE 2.7 Relative proportions of Pseudoterranova decipiens and Contracaecum osculatum species complexes in definitive hosts (seals) from the North Atlantic Ocean. It shows differences in their definitive host preferences.

Anisakid Nematodes and Host–Parasite Co-evolutionary Processes

117

Pseudoterranova species could have led to a niche subdivision, evolving by character displacement (Bullini et al., 1997; Nascetti, 1992). Another example of niche subdivision is represented by the two Antarctic members of the C. osculatum complex, that is, C. osculatum D and C. osculatum E. The two sibling species share the same definitive host, the Weddell seal, Leptonychotes weddellii, in both the Weddell and Ross Seas, of Antarctica. However, the two anisakid species appear to be adapted in this sympatric area to pelagic and benthic food webs, respectively. Larval individuals (n = 1306) of C. osculatum D and C. osculatum E, genetically identified to the species level by allozymes, were found to be heavily infecting fish species in the Ross Sea (Mattiucci and Nascetti, 2007, unpublished data). Statistically significant differences were detected in the relative proportions of the two Antarctic species of the C. osculatum complex in benthic and mesopelagic fish intermediate host species. Indeed, C. osculatum D outnumbers C. osculatum E in channichthyid fish (i.e. Chionodraco hamatus, Pagetopsis macropterus) and in bathydraconids (i.e. Gymnodraco acuticeps, Cygnodraco mawsoni); while, higher relative proportions of the infection by C. osculatum E were observed in nototheniid fish (Trematomus bernacchii, T. pennellii, T. newnesi and Notothenia neglecta) (Fig. 2.8). The last group of nototheniid fish are bottom living and feed on benthic invertebrates, such as benthic amphipods (Daniels, 1982; Foster and Montgomery, 1993; Vacchi et al., 1994), whereas channichthyid and bathydraconid fish have a pelagic/mesopelagic mode of life (Daniels, 1982). In the sibling species C. osculatum D and E, competition in the intermediate hosts could be avoided by spatial separation in the pelagic and benthic components of food webs in the Antarctic marine ecosystem. On the other hand, Klo¨ser et al. (1992) suggested that C. radiatum in Antarctic waters uses a pelagic food web. This was also confirmed by the genetic identification of larval stages of C. radiatum, at a high prevalence, in pelagic channichthyid fish (i.e. Chionodraco hamatus) (Arduino et al., 1995). Life-history differences in anisakid nematodes genetically recognized at the species level were also reported in the case of the P. decipiens species complex. In the northern hemisphere, P. bulbosa is a parasite at adult stage of the bearded seal both in the North Atlantic and Pacific Oceans, and at larval stage it occurs in the flatfishes Hippoglossoides platessoides, Reinhardtius hippoglossoides and Hippoglossus hippoglossus, in Myoxocephalus quadricornis, and, rarely, in Gadus morhua macrocephalus (see Brattey and Davidson, 1996; Bristow and Berland, 1992; Mattiucci et al., 1998). The hosts so far detected for P. bulbosa are specialized benthic species, with the exception of G. morhua which is benthopelagic (Paggi et al., 1991, 1998c). This supports the hypothesis that P. decipiens (s.l.) in the northern hemisphere has a benthic life cycle (Bristow and Berland, 1992). A similar benthic life-history pathway can be suggested for P. azarasi in Pacific

118

Simonetta Mattiucci and Giuseppe Nascetti

C. osculatum E C. osculatum D Notothenia neglecta

Bentihic

Trematomus bernacchii Trematomus pennelli

Gymnodraco acuticeps Cygnodraco mawsoni Cryodraco antarcticus Chionodraco hamatus Chaenodraco wilsoni Pagetopsis macropterus 0

10

20

30

40

50

60

70

80

90

100

Pelagic and mesopelagic

Trematomus newnesi

FIGURE 2.8 Relative proportions of larval specimens of the Antarctic anisakids Contracaecum osculatum D and C. osculatum E, genetically identified from fish hosts. It depicts differential distribution of the two species in pelagic, mesopelagic and benthic fish hosts respectively, in Antarctica.

waters. It was detected genetically at the adult stage in the benthic feeding Steller’s sea lion, Eumetopias jubatus, and also in mixed infections with P. bulbosa in the bearded seal; and at larval stage in the flatfish Hippoglossus hippoglossus (see Mattiucci et al., 1998; Paggi et al., 1998c). Benthic and benthopelagic fish species (i.e. Chaenocephalus aceratus, Notothenia coriiceps, N. neglecta, Trematomus newnesi) are found infected by the Antarctic taxon P. decipiens E (our unpublished data), confirming the suggestion of Palm et al. (1994), according to which P. decipiens (s.l.) completes its life cycles in the Weddell Sea environment in organisms of the benthic food web. Interactions in seals between anisakid species of different genera are likely to exist, since they often co-occur in the same host individual. As anisakine nematodes become numerous in seal stomachs, they may be increasingly prone to the consequences of intra- and inter-specific competition. These effects could provoke high mortality rates in one species, retard its maturity or result in lower uterine egg counts in female worms. McClelland et al. (1985) and Burt (1994) speculated that P. decipiens (s.l.) abundance in seal stomachs may be limited as a consequence of ‘competitive exclusion’ by the related nematode C. osculatum (s.l.). McClelland et al. (1985) suggested that C. osculatum (s.l.) interfered with the proliferation of P. decipiens (s.l.), since seasonal and geographical reductions in the numbers of P. decipiens (s.l.) in grey seals often coincided with increases in the abundance of C. osculatum (s.l.). This negative correlation between the abundance of the two anisakid nematodes was not supported by the data

Anisakid Nematodes and Host–Parasite Co-evolutionary Processes

119

of Brattey and Stenson (1993) according to which, these changes could be equally well due to differences in diet among grey seals, as these tend to acquire infection with larval P. decipiens (s.l.) by preying more on benthic fish (McClelland et al., 1990); whereas others feed more on pelagic fish species, such as capelin, Mallotus villosus, or cod, Gadus morhua, that harbour larvae of C. osculatum (s.l.) (see Pa`lsson and Beverly-Burton, 1984). Marcogliese (1997), remarking on the relatively small size and low uterine egg counts of female P. decipiens (s.l.) in grey seals from Anticosti Island, suggested that P. decipiens may be suppressed as a consequence of competitive pressure from C. osculatum. McClelland (2002) suggested, however, that significant negative correlation between the abundance of the two nematode species, P. decipiens and C. osculatum, could result when samples taken in a given area also include migrant seals acquiring infections in other areas. However, when members of the P. decipiens and C. osculatum complexes infest the same seal host, a differential use of resources could take place, as the former occupy the seal intestine when the stomach (which is probably the suitable microhabitat for both anisakids) is highly parasitized by C. osculatum (s.l.) (see Berland, 1964; Klo¨ser et al., 1992; Paggi and Bullini, 1994; Mattiucci and Nascetti, unpublished data). Berland (1964) speculated that C. osculatum competes with and displaces Phocascaris spp. from the stomach of harp Pagophilus groenlandicus, and hooded seal Cystophora cristata. This result was later confirmed by genetic identification (allozymes) of several individuals of C. osculatum B and Phocascaris spp. It is likely that C. osculatum B is a better competitor than Phocascaris phocae or P. cystophorae during syntopic occurrence, at the adult stage, in the same individual seal host. Marked differences were indeed observed in their relative proportions when they occur syntopically in the harp and common seals. In these cases, the stomach is mostly occupied by C. osculatum B, while P. phocae and P. cystophorae are displaced to the upper intestine. In contrast, the hooded seal has always been found parasitized solely by P. cystophorae/P. phocae in both sites (stomach and intestine). In this seal, Phocascaris spp., in the absence of a better competitor, usually occupies the stomach. These phenomena of niche subdivision, promoting character displacement, could reduce the inter-specific competition between two closely related anisakid species (i.e. C. osculatum B and P. phocae/P. cystophorae) for a trophic resource (Nascetti, 1992; Paggi and Bullini, 1994).

8. ANISAKIDS AS BIOLOGICAL INDICATORS In recent years, marine parasites have been widely used in biological and ecological surveys of marine ecosystems as biological indicators of food chain structure (Thompson et al., 2005), pollution (Sures, 2004), global

120

Simonetta Mattiucci and Giuseppe Nascetti

climate changes (Brooks and Hoberg, 2007; Poulin, 2006), anthropogenic impacts and environmental stresses (Marcogliese, 2005), fish stock assessment (MacKenzie, 2002) and of general ‘ecosystem health’ (Marcogliese, 2005). Anisakid parasites are an integral part of aquatic ecosystems. They are indeed important components of any ecosystem. Not only do they play key roles in population dynamics and community structure, they can provide important information about the general biodiversity of the ecosystem. Recent data on the possible use of anisakid nematodes are presented in this section as biological indicators of (i) the definition of fish stocks within a multidisciplinary approach, (ii) integrity and stability of trophic webs and (iii) habitat disturbance.

8.1. Anisakis spp. larvae as biological tags of fish stocks One of the major goals of the last decade in the study of the parasite fauna of aquatic organisms has been, among the others, the assessment of fish stocks, their movement and recruitment. Indeed, the use of parasites to discriminate fish host populations has been one of the most useful approaches in a multidisciplinary study of fish stock detection and characterization. Nowadays, the modern concept of ‘fish stock’ integrates all the information gathered from a broad spectrum of techniques, ranging from morphology (morphometrics, meristic, otolith microchemistry) to biology (life-history characteristics, mark-recapture, parasites) and genetic structure (allozymes, microsatellites, DNA sequences) of the fish hosts throughout their geographical range. In this ‘holistic approach’ to the definition of fish stock (Begg and Waldman, 1999), among the biological approaches, the use of parasites as ‘biological tags’ has become a useful tool, mainly concerning species with a high commercial value in fisheries. Trends in the use of parasites as biological tags in population studies of marine fish species was recently reviewed (MacKenzie, 2002). Among the parasite species that have been used in fish stock definition, the larval anisakid nematodes of the genus Anisakis represent one of the best biological tags (MacKenzie, 2002). Allozyme markers were first used to differentiate populations of herring, Clupea harengus L., in the North Atlantic, on the basis of different allele frequencies observed at one enzymatic locus in Anisakis larvae (Beverly-Burton et al., 1977). The application of diagnostic genetic markers in the identification of larval Anisakis to species level, and the biogeographical aspects of Anisakis spp., have, in recent years, allowed the fish stock identification of several demersal and pelagic species (Mattiucci et al., 2007a). A parasitological survey carried out on several fish specimens of the European hake, Merluccius merluccius, from 14 localities in the North East Atlantic and

Anisakid Nematodes and Host–Parasite Co-evolutionary Processes

121

Mediterranean Sea included in the range of distribution of the fish species, has allowed detection, by allozyme markers, of the occurrence of different biological species of larval Anisakis (n = 1950). Indeed, seven species of Anisakis were genetically recognized [A. pegreffii, A. simplex (s.s.), A. typica, A ziphidarum, A. physeteris, A. brevispiculata and A. paggiae] showing statistically significant differences in their distribution in the hake sampled from Mediterranean and Atlantic waters (Mattiucci et al., 2004) (Fig. 2.9). This indicated that (i) two stocks of M. merluccius from Mediterranean and Atlantic waters exist; (ii) in the North Atlantic area, at least two different population units of European hake are present; one northern and another from off the Atlantic coast of Morocco; and (iii) some sub-structuring in the Mediterranean population was also recognized (Mattiucci et al., 2004). These findings are in accordance with genetic results obtained from allozyme markers on the same fish host samples, demonstrating that Atlantic and Mediterranean populations of M. merluccius belong to two different stocks. These two populations are separated at the Almeria-Oran front (Cimmaruta et al., 2005), as this front has been proved to be the boundary between Atlantic and Mediterranean stocks of several marine organisms (Pannacciulli et al., 1997; Rios et al., 2002). In this area a strong inflow transports surface Atlantic waters into the Albora´n Sea, creating an Atlantic domain within the Mediterranean, and a partial isolation of the Albora´n with respect to the other Mediterranean water masses. It was also demonstrated that salinity and temperature are responsible for maintaining the genetic differentiation between the two fish population groups through selective processes (Cimmaruta et al., 2005). In the case of horse mackerel, Trachurus trachurus, different populations can be distinguished according to the spatial distribution of the two most dominant parasite species, genetically recognized as A. simplex (s.s.) and A. pegreffii. Indeed, the significant differences indicated by their relative proportions between the samples from the Mediterranean Sea (excluding the Albora´n Sea) in comparison with the Atlantic ones seem to indicate the discrete sub-structuring of populations of T. trachurus in Mediterranean and Atlantic waters. Moreover, in Atlantic waters, the differences found in the proportion of A. pegreffii between the southern horse mackerel samples and those remaining seem to suggest a possible existence of a ‘southern’ sub-population of T. trachurus readily distinguished from a ‘northern’ one. The samples of T. trachurus from the Albora´n Sea were closer to the southern populations of the Atlantic than to the other samples from the Mediterranean Sea. This finding, and particularly the presence of A. simplex (s.s.) in some individual hosts from the Albora´n Sea, suggests that these populations could be the result of a migration of this small pelagic fish species from Atlantic waters into this extreme western area of the Mediterranean Sea, and also possibly

122

Simonetta Mattiucci and Giuseppe Nascetti

A

Merluccius merluccius

B

Trachurus trachurus

C

Xiphias gladius

A. simplex (s.s.) A. pegreffii A. brevispiculata A. physeteris A. ziphidarum A. typica A. paggiae Anisakis sp.

50 45 40 35 –50 –45 –40 –35 –30 –25 –20 –15

50 45 40 35 –50 –45 –40 –35 –30 –25 –20 –15 –10 15 10 5 0 –30

–25

–20

–15

–10

–5

FIGURE 2.9 Relative proportions of larval specimens of Anisakis spp. from different fish hosts: (A) Merluccius merluccius; (B) Trachurus trachurus; (C) Xiphias gladius throughout their range of distribution. It shows the use of biogeographical aspects of Anisakis spp., genetically identified to species level, as biological tags for their fish stocks definition in European waters.

Anisakid Nematodes and Host–Parasite Co-evolutionary Processes

123

mixing with Mediterranean populations. This, in turn, is related to the possible Almeria–Albora´n boundary in Mediterranean waters. Furthermore, the high percentage of fish exhibiting infection by A. simplex (s.s.) from this area was in accordance with the high dispersal capacity of this small pelagic fish (Mattiucci et al., 2008a) (Fig. 2.9). This is in agreement with the high genetic variability values shown for T. trachurus, at the nuclear level (allozymes), in the same fish samples (Cimmaruta et al., 2008). However, in contrast with the parasitological findings, a substantial genetic homogeneity, inferred from allozyme markers on the same fish samples, was found between the Atlantic and Mediterranean populations of horse-mackerel; a partial sub-structuring of T. trachurus observed between the far eastern and western Mediterranean Sea has been reported (Cimmaruta et al., 2008). As for the swordfish, Xiphias gladius, it has been recently demonstrated, using different molecular genetic markers and biological data, that the Mediterranean population is separated from the Atlantic one (Cimmaruta et al., 2006; Kotoulas et al., 2006; Lu et al., 2006; Reeb and Block, 2006). The existence of a further separation between northern and southern populations of swordfish in the Atlantic has been also suggested (Alvarado-Bremer et al., 2006). The occurrence of different species of Anisakis spp. larvae in Mediterranean and Atlantic swordfish populations was recently shown (Mattiucci et al., 2007a). In particular, the absence of A. simplex (s.s.) in the Mediterranean swordfish samples seems to support the idea that the Mediterranean population of X. gladius is separated from the Atlantic one. Moreover, in Atlantic waters, the possible existence of two populations, one northern and one southern, has been suggested (Mattiucci et al., 2007a) (Fig. 2.9). This is also supported by population genetics data obtained from the swordfish sampled in this area, indicating a possible boundary of the two stocks between 0 N and 5 S in Atlantic waters (Alvarado-Bremer et al., 2006). Recently, it has been shown that the polymorphism observed in mtDNA cox1 gene sequences, analysed for several samples of A. simplex (s.s.) collected between and within spawning seasons of its fish host (the herring, Clupea harengus), exhibited a high degree of temporal stability (Cross et al., 2007). This underlines the potential use of this species as a biological tag, and mtDNA cox1 as a suitable genetic marker for future investigation in fish host stock structure definition (Cross et al., 2007).

8.2. Anisakids as indicators of trophic web stability and habitat disturbance of marine ecosystems Food webs are networks of trophic relationships which map the location of energy flow in a community. Despite the increasing details that have been incorporated into marine food webs, some functional groups remain

124

Simonetta Mattiucci and Giuseppe Nascetti

neglected. Among these are the parasites and host–parasite interactions that are known to affect, among others, community structure and trophic relationships (Marcogliese, 2001, 2005; Poulin and Morand, 2004). The transmission pathways of parasites with indirect life cycles are fully included in food webs of aquatic ecosystems. In other words, just as food webs have exerted strong selective pressure on the evolution of parasite transmission strategies, parasites are now shaping some of the ecological properties of existing food webs. The transmission routes of anisakid nematodes follow closely the trophic relationships among their successive hosts, and thus they are parasites embedded in food webs. As a consequence, the completion of life cycles as complicated as those of anisakid nematodes requires a stable trophic web. As a result, the life cycle of anisakid nematodes in marine ecosystems with various degrees of habitat disturbance could be affected by changes in host population size. Indeed, when the population size of the hosts participating in the life cycle of these parasites is reduced due to different causes (pollution, by-catch of marine mammals, viral diseases of marine mammals, overfishing, etc.) (Fig. 2.10), the population size of their anisakid endo-parasites could also be reduced. This would result in a higher probability of genetic drift in the parasite gene pools and, consequently, a decrease in their genetic variability values. It has recently been shown (Mattiucci and Nascetti, 2007) that the distribution of the genetic variability of anisakid nematode populations in geographical areas with different levels of environmental stress is likely to reflect the influence of a range of factors that could promote their genetic diversity. These include a large effective parasite population size, the wide range, availability, and population size of their hosts, and the stability of marine trophic webs. The values of the genetic variability [estimated at the parameters of percentage of polymorphic loci (P), mean number of alleles per locus (A) and expected heterozygosity per locus (He)], obtained at 19 allozyme loci, were compared among 53 populations of anisakid nematodes belonging to 20 species of Anisakis, Pseudoterranova and Contracaecum from several hosts in the Boreal and Austral regions (Mattiucci and Nascetti, 2007). Austral populations of species belonging to these three genera exhibited significantly higher genetic variability values than those from the Boreal regions [expected mean of heterozygosity per locus, He = 0.19 (in Austral populations) and He = 0.09 (in Boreal populations) (P < 0.01)] (Fig. 2.11). A more remarkable difference in their genetic variability values was observed when only Antarctic and subAntarctic populations were compared directly with Arctic and sub-Arctic populations [He = 0.23 and He = 0.07 (P < 0.001), respectively] (Mattiucci and Nascetti, 2007). One conclusion is that the observed values of genetic variability could be related to extreme latitudes, a parameter often considered as relevant (Nevo et al., 1984). However, the data suggested that a

Anisakid Nematodes and Host–Parasite Co-evolutionary Processes

Host

Definitive

Intermediate/ paratenic

First intermediate

125

Habitat disturbance By-catch and hunting, viral diseases

Overfishing, contaminants

Global warming, sea water acidification, pollution

FIGURE 2.10 Relationship between habitat disturbance and host population size. It depicts causes of habitat disturbance (right) that can affect the population size of definitive and intermediate hosts (at different trophic levels of a marine food web) involved in speculative life-cycle pathways of anisakid nematodes of the genera Anisakis, Pseudoterranova and Contracaecum (includes only species from pinnipeds). (Modified from Mattiucci and Nascetti, 2007, with permission from Elsevier.)

significantly higher level of genetic variability found in the Antarctic members considered (i.e. C. osculatum D, C. osculatum E, C. radiatum, P. decipiens E, A. simplex C and A. pegreffii populations from sub-Antarctic regions) coincide with a lower degree of habitat disturbance (e.g. overfishing, by-catch of cetaceans, hunting and diseases mortality of seals, sea water pollution and acidification). This would allow host species to reach higher population sizes, resulting in higher anisakid population sizes, with a reduced probability of genetic drift phenomena in the parasite gene pools. Consequently, a higher level of genetic diversity in the Antarctic populations of these nematodes was observed. Likewise, a much higher abundance and intensity of infection was observed in the Antarctic and sub-Antarctic populations and species of anisakid nematodes, where more than 105 individual worms were typically collected from a single host, compared to