Adventures with poxviruses of vertebrates


*Tel.: +61 (2) 62492526; Fax: +61 (2) 62484723, E-mail address:


Because they were the largest of all viruses and could be visualised with a light microscope, the poxviruses were the first viruses to be intensively studied in the laboratory. It was clear from an early date that they caused important diseases of humans and their domestic animals, such as smallpox, cowpox, camelpox, sheeppox, fowlpox and goatpox. This essay recounts some of the early history of their recognition and classification and then expands on aspects of research on poxviruses in which the author has been involved. Studies on the best-known genus, Orthopoxvirus, relate to the use of infectious ectromelia of mice as a model for smallpox, embracing both experimental epidemiology and pathogenesis, studies on the genetics of vaccinia virus and the problem of non-genetic reactivation (previously termed ‘transformation’) and the campaign for the global eradication of smallpox. The other group of poxviruses described here, the genus Leporipoxvirus, came to prominence when the myxoma virus was used for the biological control of Australian wild rabbits. This provided a unique natural experiment on the coevolution of a virus and its host. Future research will include further studies of the many immunomodulatory genes found in all poxviruses of vertebrates, since these provide clues about the workings of the immune system and how viruses have evolved to evade it. Some of the many recombinant poxvirus constructs currently being studied may come into use as vaccines or for immunocontraception. A field that warrants study but will probably remain neglected is the natural history of skunkpox, raccoonpox, taterapox, yabapox, tanapox and other little-known poxviruses. A dismal prospect is the possible use of smallpox virus for bioterrorism.


In accordance with the request from the editors, this essay will concentrate on the two fields of virology on which my work has focused over the last 50 years: the orthopoxviruses, principally ectromelia, vaccinia and variola viruses, and the leporipoxviruses, principally myxoma virus. With colleagues as co-authors, I have written books on both of these topics [1–3]; in this essay I will try to summarise the ‘ancient history’ of the subject, to extract the highlights of work with which I have had some connection, and briefly to look into the unknown future to see what it may hold.

2Early history of poxviruses

Smallpox, caused by an orthopoxvirus, was once the most serious disease of humankind. Unlike malaria, it was not limited by climate, and unlike plague, it was always present. The agent that caused it was the first virus to be seen with a microscope [4], it provided the first example of inoculation with a virus as a preventative measure against the disease caused by that agent [5], it was the first disease against which an effective vaccine was used [6], the group to which it belonged was the first to be correctly classified as what came to be known as a family [7] and it was the first human disease to be eradicated globally [5].

The virus used to prevent smallpox (vaccinia virus) became a model for early biological and biochemical studies of viruses [8]. It was the first animal virus to be purified sufficiently to show that it contained DNA but not RNA [9] and to be visualised by electron microscopy [10]. It also provided an early model for assay methods comparable with colony counts for bacteria, namely an assay for infectivity by pock counts on the chorioallantoic membrane of the developing chick embryo [11,12]. More recently, it was the first virus to be used as a vector for carrying foreign genes into animals in such a way that the proteins for which they coded were expressed [13–15].

The model virus of the other genus on which this review will focus, Leporipoxvirus, also has a history going back to the early days of virology. In 1893 Guiseppe Sanarelli, an Italian bacteriologist who had worked in the Pasteur Institute in Paris, went to Montevideo in Uruguay to establish an Institute of Experimental Hygiene. In 1896 a mysterious and lethal infectious disease broke out among his laboratory rabbits, which had been imported from Brazil, from which he recovered a non-cultivable and invisible infectious agent which he named the ‘myxomatogene virus’[16]. It shared with foot-and-mouth disease virus the distinction of being the first virus of vertebrates (in the modern sense of ‘virus’) to be described as such. Until the mid-1930s myxomatosis excited little interest except among people in Brazil and California who were trying to farm European rabbits, although the Brazilian microbiologist H.B. Aragão (quoted in [3]) suggested in 1918 that it might be used to control pest rabbits in Australia. In later studies Aragão went on to show that the myxoma virus was morphologically similar to the variola virus [17], that its natural host was the Brazilian forest rabbit and that it could be transmitted mechanically by mosquitoes and fleas [18]. It was eventually imported into Australia in the mid-1930s and, after laboratory and field studies that extended over several years, it was released among Australian wild rabbits in the early 1950s. Two years after its spread in Australia it was illegally introduced in France and spread widely in Europe, to the dismay of rabbit farmers and hunters but the delight of foresters and other farmers [3]. It proved to be the most effective biological control method for a vertebrate pest that has ever been discovered.

3Classification of poxviruses

The earliest classifications of viruses were based on disease symptoms. Certain diseases of humans, cow, sheep, horse and pig were classed as ‘poxes’ because disease with which they were associated were characterised by pocks on the skin. As it turned out, several of these diseases were indeed caused by poxviruses, but the deficiencies of a classification based on clinical symptoms was highlighted by the inclusion of chickenpox and the ‘great pox’ (syphilis) in the same group as the smallpox virus. Of course, the fallacies of such a method of classification of viruses (as distinct from diseases) are even more apparent with the terms ‘hepatitis viruses’, ‘encephalitis viruses’ and ‘haemorrhagic fever viruses’.

Being the largest of all animal viruses and visible in stained smears by light microscopy, the poxviruses were the first ‘group’ of viruses to be described, i.e. viruses that were not serologically related but appeared to have a more general resemblance, in size and certain other characteristics. As it turned out, this seemingly superficial resemblance has proved to be remarkably robust in terms of the characteristics now used to classify viruses. Aragão [17] pointed out the resemblance between the viruses of ‘variola, myxoma, bird-epithelioma, molluscum contagiosum, etc.’ and in 1933 Goodpasture [19] formally proposed that vaccinia-variola, fowlpox, horsepox, sheeppox, goatpox, swinepox and molluscum contagiosum viruses should be grouped together as the genus Borreliota.

Viral classification took a step forward in 1948 when Holmes (a plant virologist) published a comprehensive classification of viruses in the standard book on the classification of bacteria [20]. Technically, it was a step backwards, because it placed major reliance on the symptomatology of disease. However, this ‘nomenclatural bombshell’ stimulated others to study viral classification seriously and after several discussion papers [21,22] in which Sir Christopher Andrewes (then the leading virologist in the UK) figured prominently, the Fifth International Congress for Microbiology in 1950 gave serious consideration to questions of viral classification and nomenclature. These were followed up at the next Congress, in 1953, by the establishment of study groups to consider five groups of viruses, of which the poxviruses were one. On behalf of the Poxvirus Study Group, in 1957 Fenner and Burnet [23] published a short description of the poxviruses of vertebrates (which are the only ones to be discussed in this essay) that has remained the basis of subsequent classification in respect of the criteria used and subdivisions adopted. The basic features of the family are the large brick-shaped or ovoid virions, with a genome consisting of a single linear molecule of covalently closed, double-stranded DNA, between 130 and 220 kb in length. Unlike most other DNA viruses, poxviruses replicate in the cytoplasm of the cell. The family Poxviridae is subdivided into two subfamilies, Chordopoxvirinae and Entomopoxvirinae, found in vertebrates and insects, respectively. Viruses of the subfamily Chordopoxvirinae are subdivided into eight genera, distinguished from each other primarily by serologic cross-reaction and cross-protection (Table 1).

Table 1.  Classification of the poxviruses of vertebrates
Family: Poxviridae; subfamily: Chordopoxvirinae
Genusprototype virus
Orthopoxvirusvaccinia virus
Parapoxviruspseudocowpox virus
Capripoxvirussheep pox virus
Suipoxvirusswinepox virus
Leporipoxvirusmyxoma virus
Avipoxvirusfowlpox virus
Molluscipoxvirusmolluscum contagiosum virus

4The genus Orthopoxvirus

The genus Orthopoxvirus, with which the first section of this essay is concerned, contains 10 species (Table 2). The four species with which I have worked will be discussed: ectromelia, vaccinia, variola and monkeypox viruses.

Table 2.  Species of the genus Orthopoxvirus
  1. aOriginal host unknown. Buffalopox and rabbitpox have been found as enzootic infections in buffaloes and rabbits, but originated from smallpox vaccine virus.

SpeciesHost range in laboratory animalsAnimals found naturally infectedGeographic range of natural infections
Camelpox virusnarrowcamelsAfrica and Asia
Cowpox virusbroadnumerous: carnivores, cow, elephant, humans, rats; natural hosts: gerbils, other rodentsEurope and former USSR
Ectromelia virusnarrowmice; natural host unknown, possibly volesEurope
Monkeypox virusbroadapes, monkeys, squirrels, humans; natural hosts, squirrelswestern and central Africa
Raccoonpox virus(?) broadraccoonUSA
Tatera poxvirusnarrowTatera kempi (a gerbil)Western Africa
Uasin Gishu poxvirusmediumhorse (natural host unknown)Kenya, Zambia
Vaccinia virus (smallpox vaccine virus)broadnumerous: buffalo, cow, man, pig, rabbitaIndia (buffalopox); Europe and USA (rabbitpox)
Variola virusnarrowhumans (now eradicated)formerly worldwide
Vole poxvirus(?) broadvolesUSA

5Studies with ectromelia virus

Because of the absence of work with the ectromelia virus in many countries it is convenient to summarise this account in three time periods: early and medium-term, which are almost restricted to work done in Australia, and the modern era, during which the advantages in the mouse as a model virus-host system are being increasingly exploited.

5.1Early work (1930–1948)

The ectromelia virus was discovered in 1930 by Marchal [24]. Being a virus infection that was naturally transmitted from one mouse to another, it was immediately used by Greenwood et al. [25] to expand their long-term experiments on experimental epidemiology, previously restricted to the bacterial diseases, mouse typhoid and mouse pasteurellosis.

After the discovery of haemagglutination by influenza virus [26], Burnet made this the main focus of virological work in the Walter and Eliza Hall Institute [27]. Being a ‘collector’ by nature (beetles in his childhood), he examined as many viruses as were available in Australia to see how common this phenomenon was. In addition to showing that Newcastle disease and mumps viruses, like influenza virus, caused haemagglutination and subsequently eluted from the red cells [28,29], he found that ectromelia and vaccinia viruses caused a rather different kind of haemagglutination and did not elute. Serological and cross-protection tests then showed that ectromelia virus was a member of the vaccinia-variola virus group, i.e. an Orthopoxvirus[30].

In 1945 Burnet, who had just been appointed as Director of the Walter and Eliza Hall Institute, was interested in extending Greenwood and Topley's studies on experimental epidemiology. At this time I was working as a malariologist in the Australian Army Medical Corps in the Halmaheras and when I was discharged, early in 1946, Burnet invited me on to work on the experimental epidemiology of infectious ectromelia [31]. He had chosen ectromelia as the model virus for such studies because it belonged to the same group as variola virus. Ectromelia was an unexplored scientific goldmine. Although they had had no trouble with the bacterial pathogens, Greenwood et al. [25] had experienced trouble with contamination of mice with ectromelia virus in ‘clean’ cages and in general ectromelia virus was regarded as dangerous to work with because of laboratory cross-infection. Such events had been common in Europe, China and Japan, where breeding colonies often harboured unrecognised infection [32]. Disastrous outbreaks of the disease among laboratory mice in the USA following the importation of mice from Europe had led to a ban on its deliberate use there [33]. Fortunately, I had no such trouble, probably because the work was conducted in a special room and the only persons who handled the mice and the cages were myself and my assistant and we were careful to wash our hands after each cage.

The first experiments were done to elucidate the natural history of ectromelia; natural transmission, immunity (after recovery or after vaccination with vaccinia virus and via the placenta or in milk), the virulence of different strains of virus and the effect of age on the response [34]. I also carried out experiments on both continuing and closed epidemics along the lines that had been developed by Topley, among which were studies of the efficacy of vaccination with vaccinia virus [35]. For many years these experimental epidemics attracted little attention, but in the late 1970s, when Anderson and May began their studies on modelling epidemics [36], these experiments [37] and those of Greenwood and his colleagues [25] provided them with the only available series of carefully studied, long-continued, experimental epidemics.

As is so often the case in experimental science, the most important discovery with ectromelia virus was an unexpected by-product of the work on experimental epidemics. To put what follows in context, most outbred mice (and nothing else was available in Australia at the time) die of acute hepatitis after being infected with ectromelia virus. A few survive, often with an amputated foot, which was the reason the Marchal proposed the name ‘infectious ectromelia’ for the newly discovered disease. In the course of the epidemiological experiments we noticed that the mice that did not die of acute hepatitis developed a rash, usually just after most other mice had died [38]. Here then was a disease caused by a virus closely related to variola virus, in that most convenient of all laboratory animals, the mouse, that like smallpox was characterised by a pustular rash if the mice lived long enough. Experiments therefore took a new turn, namely to use this model to find out what happened during the long incubation period of smallpox and, by analogy, that of other generalised exanthematous diseases [39,40] We now called the disease mousepox, but continued to use the term ‘ectromelia virus’ just as the virus that causes smallpox is usually called variola virus. The Lancet paper [40] was reproduced as a ‘classic’ nearly 50 years later [41].

5.2Later studies in Australia (1958–1988)

I gave up experimental work with ectromelia virus when I left the Walter and Eliza Hall Institute in 1948 for a period of study at the Rockefeller Institute of Medical Research. I had asked whether I could take ectromelia virus with me to carry out studies for which we did not have the apparatus in Australia. The suggestion was met with horror and importation of ectromelia virus into the USA was formally banned in 1954 [42]. However, it has been and is still being used very effectively in the John Curtin School of Medical Research as a model virus.

In 1956, Mims [43] began investigations of the pathogenesis of a number of viral infections in mice by supplementing infectivity titrations and histology by the use of fluorescent antibody staining. Among the poxviruses, he investigated ectromelia and cowpox viruses, looking at the role of macrophages in determining susceptibility, cell-associated viraemia and the role of the capillary endothelium, especially in relation to clearance by macrophages. He noted the neglect of such studies elsewhere, ascribing it to ‘the boom in tissue culture’. In a later review [44] he summarised his results on the pathogenesis of rashes in viral diseases, again noting the value of mousepox as a model and of fluorescent antibody staining as a method of investigating pathogenesis at the cellular level.

Blanden, an Adelaide dental graduate, had been studying the cellular response to infection with the bacterium Listeria monocytogenes at the University of Adelaide with Mackaness [45] and in 1965 he moved with Mackaness to the Trudeau Institute for Medical Research in upper New York State. In 1968 he enrolled for a Ph.D. degree in the John Curtin School and applied his knowledge of cell-mediated immunity to the study of mousepox. In a series of classical papers (reviewed in [46]), he demonstrated the importance of cell-mediated immunity, compared with serum antibody, in recovery and protection in generalised viral infections. His work set the stage for work by Doherty and Zinkernagel [47], in the same Department, on MHC restriction.

5.3The modern era

The modern era can be said to have begun when study of ectromelia virus in laboratories in the USA became acceptable. In 1980 I formally retired from the Australian National University. My first post-retirement job was a recall to my first love as a virologist, mousepox. In 1979–1980 several outbreaks of mousepox initiated by importation of mice from laboratories in Europe had occurred in valuable colonies of inbred mice in several places in the USA, including the National Institutes of Health. A major conference to discuss the problem was organised by the National Institutes of Health [33]. As one of the few virologists then available who had studied the disease, I was asked to provide a background paper [48]. One consequence of the conference was the provision of a special high security laboratory at the National Institutes of Health to study ectromelia virus, which produced valuable data on the resistance of different pure-bred strains of mice and identified several ‘resistance’ genes [49].

These results and the great advantage of using a disease which was a transmissible infection in the mouse, the best of all experimental animals as a model, led to several scientists in the USA and the UK to join with Australian scientists in using the mousepox model for studies of immunoevasion and immunomodulation [50–52].

6Genetic studies with vaccinia virus

Although after 1948 I did not study ectromelia again, in 1957 I was attracted to another orthopoxvirus, vaccinia virus, as a tool for trying to unravel the virulence of poxviruses, which had been such a striking feature of myxoma virus, the agent with which I had been working since 1951 (see below). Myxomatosis in the wild European rabbit, in both Australia and Europe, provided a wonderful natural experiment on the coevolution of a virus and its host. I was intrigued by the extreme virulence of the original strains of myxoma virus and the rapid selection for strains of somewhat reduced virulence, but myxoma virus was not a good virus for laboratory work. It grew poorly in cell cultures and the only test for virulence was the inoculation of laboratory rabbits, which were large and expensive animals, compared with mice. Vaccinia virus looked much more promising for the study of the genetics of virulence in poxviruses, which has turned out to be a much more complex problem than I had imagined and more complex than many present-day virologists realise.

To decide what strains would be most suitable for experiments on poxvirus genetics, I made a survey of 24 strains of vaccinia and cowpox viruses from different laboratories all over the world [53]. I selected two strains of vaccinia virus that had strongly contrasting marker characters, vaccinia 7N and rabbitpox Utrecht, and showed that genetic recombination occurred between them [54,55]. In those days Australian scientists did not travel abroad often, but in 1957 I spent four months visiting laboratories all over the USA and Europe. I went to the University of Illinois at Urbana to visit Salvador Luria and discuss my genetic studies. He remarked that judging from phage genetics I would not get very far if I used two different strains; I should try to use mutants of one strain. Fortunately, a suitable strain was at hand, for I had already noticed that both cowpox virus and rabbitpox virus (a variant of vaccinia virus [56]) produced pocks on the chorioallantoic membrane with a haemorrhagic centre, but with both viruses about 1% of the pocks were non-haemorrhagic (white) [53,57]. Further, these white pocks differed from each other in appearance and the causative viruses often differed in other characteristics, such as plaque morphology, haemagglutinin production and virulence for the mouse and the rabbit. Since on passage they ‘bred true’, it was reasonable to regard them as stable mutants. When cells were infected with a mix of two different white pock mutants, a proportion of the progeny produced wild-type haemorrhagic pocks. Using data from a number of such crosses, Gemmell and Cairns [58] developed a primitive genetic map of rabbitpox virus. Later I studied the characteristics of host range (PK cell) mutants of rabbitpox virus [59]. It turned out that all PK mutants produced white pocks, but not vice versa. These mutants subsequently proved very useful for early molecular studies of poxvirus genetics [60].

From my earlier work with myxoma virus I had become aware of an unexplained phenomenon that had been discovered in 1936, during experiments with myxoma virus and the related leporipoxvirus, fibroma virus. Stimulated by Griffith's work on pneumococcal transformation [61], Berry and Dedrick [62] inoculated a mixture of heat-killed myxoma virus and active fibroma virus into rabbits. Some of the rabbits died of myxomatosis; by analogy with Griffith's studies they called the phenomenon the ‘transformation’ of fibroma virus into myxoma virus. In 1952 I had tried unsuccessfully to reproduce this phenomenon by inoculating mice with a mixture of heat-inactivated ectromelia virus and active cowpox virus. However, following the initial work on recombination using two different strains of vaccinia virus [55], we demonstrated that heat-inactivated rabbitpox virus could be ‘transformed’ (to use the Berry–Dedrick term) by several other poxviruses, belonging to different genera, namely cowpox and ectromelia (orthopoxviruses), myxoma and fibroma (leporipoxviruses) and fowlpox virus (an avipoxvirus), but not by viruses belonging to different families [63]. Moreover, the ‘transformed’ rabbitpox virus invariably resembled the virus that had been inactivated in all of the five characters that could be tested. We therefore chose to call the phenomenon ‘non-genetic reactivation’, in contrast to the well-known genetic processes of multiplicity reactivation and marker rescue. The mechanism was elucidated by W.K. Joklik [64], who showed that a viral protein (the ‘inducer protein’, which is inactivated by heating) is released from the active poxvirus during the first stage of uncoating and stimulates the synthesis by the host cell of a protein which releases the viral DNA from the cores of both heated and active virus particles.

7The eradication of smallpox

In 1958 it was proposed by the representative of the USSR and accepted by the World Health Assembly that a campaign should be launched to eradicate smallpox world-wide, based on the vaccination of at least 80% of the population of endemic countries. This had some success, especially in smaller countries, but by 1967 it was clear that this approach was not going to succeed in large endemic countries like those of the Indian subcontinent. The World Health Organisation therefore decided to launch an Intensified Smallpox Eradication Programme and a small Smallpox Eradication Unit, headed by an American, D.A. Henderson, was established at WHO Headquarters in Geneva. Its program was based on childhood vaccination plus a well-organised campaign of surveillance and containment in all endemic countries. Henderson also insisted on the need for continuing research on problems that arose during the campaign.

I was a member of WHO's Advisory Panel on Virus Diseases and in 1969 I was asked to participate in the first meeting of a panel of experts on poxviruses to discuss the possible existence of an animal reservoir of smallpox, a matter of great concern to the Smallpox Eradication Unit. The meeting was asked to assess the significance of monkeypox virus, which had been isolated from laboratory monkeys in Copenhagen in 1958 [65] but not, at that time, from humans. The next year, 1970, monkeypox virus was isolated from several cases of a smallpox-like disease in West Africa and Zaire [66]. The work of this committee soon became much more important, because of reports by virologists from Moscow, beginning in 1971, that they had isolated what was called ‘whitepox virus’ from four different species of forest animals in Zaire (reviewed in [5]). The significance of this report was that ‘whitepox virus’ was indistinguishable from variola virus. In 1978 the Russian workers reported that they had discovered the origin of the whitepox strains, claiming that they were white pock mutants of monkeypox virus [67]. My experience with white pock mutants of rabbitpox virus convinced me that this claim was wrong, because all their so-called white pock mutants of monkeypox virus were identical and indistinguishable from variola virus, whereas every rabbitpox virus white pock mutant that I had isolated was different. It therefore seemed likely that these ‘monkeypox virus white pock mutants’ were due to laboratory contamination with variola virus. Further, if they had isolated variola virus from their laboratory stocks of monkeypox virus, was it not likely that the ‘whitepox’ viruses were also laboratory contaminants? The question was of the utmost importance, for if there was a wildlife reservoir of variola virus smallpox eradication was impossible. The probability of laboratory contamination was later shown to be overwhelming and this conclusion was eventually accepted by the Moscow workers [5].

In 1977, as the smallpox eradication campaign approached a successful conclusion, it became important to extend and formalise the certification of eradication, for unless it could be conclusively demonstrated that smallpox had been eradicated, the benefits of eradication would not be realised. I had been involved with the Programme's scientific committee since 1969 and was chairman of its meeting in Geneva in 1976. In 1977 I was asked to serve as Chairman of a Consultation on the Worldwide Certification of Smallpox Eradication, a large committee of senior public health officials and virologists that met in Geneva in October that year and mapped out a strategy for global certification. Subsequently I was appointed chairman of the Global Commission for the Certification of Smallpox Eradication, which met in Geneva in October 1978 and December 1979. At the last meeting, on Sunday 9 December 1979, all members signed a declaration affirming that the global eradication of smallpox had been achieved and agreed to 19 recommendations for actions relating to public health and other matters in the post-smallpox world [68]. In May 1980 the report and all of its recommendations were accepted by the World Health Assembly.

As a follow-up to the Global Commission, WHO appointed a Committee on Orthopoxvirus Infections to oversee implementation of the recommendations of the Report of the Global Commission and I was asked to serve as chairman of this committee. Recommendation 16 of the Global Commission's report read: “WHO should ensure that appropriate publications are produced describing smallpox and its eradication and the principles and methods that are applicable to other programmes.” The outcome of this recommendation was produced 8 years later, a massive book of 1460 pages [5].

WHO had decided that the Committee on Orthopoxvirus Infections should have a lifespan of 5 years. It met regularly between 1980 and 1984, overseeing such matters as the progress of the book, the condition of the vaccine reserve and the destruction of stocks of variola virus in all laboratories except the two WHO Collaborating Laboratories on Smallpox and Related Infections in Atlanta, USA, and Moscow, USSR. Then, officially, it ceased to exist, but in 1986 it was recalled as an ad hoc Committee, to consider whether the stocks of variola virus in Atlanta and Moscow (thought to be the only stocks in the world) should be destroyed. The Committee decided that they should be, but agreed to a suggestion that a few representative strains should first be sequenced. This was done and the ad hoc Committee met again, and again recommended destruction. However, because of political intervention by some powerful States, the item was withdrawn from the agenda. Eventually, in 1996 it came before the World Health Assembly and it was agreed that they should be destroyed in June 1999. In May 1999 the destruction was again deferred, until 2002. The early delays appear to have been due to lobbying by defence authorities of some of the industrialised countries, probably because they had advanced knowledge, kept secret, of the biological weapons project of the Soviet Union.

8Future research on orthopoxviruses

Forecasts of future discoveries inevitably omit those that are the most interesting, i.e. those that are completely unexpected and open up new vistas. Given this omission, there are several obvious gaps in even the most elementary knowledge of several orthopoxviruses. Raccoonpox, skunkpox, taterapox, Uasin Gishu and volepox viruses (Table 1) have never been adequately studied in the laboratory or in the field. Virtually no molecular studies have been made of camelpox virus, although in restriction patterns and pock morphology it is rather similar to variola virus. These deficiencies should be repaired, but they are unlikely to attract funding.

The early work on using vaccinia or fowlpox virus as a vector for foreign genes has advanced greatly in recent years and there is no doubt that this work will continue to expand. The discovery that a much more effective responses to DNA vaccines can be obtained if these are followed up with a suitable poxvirus recombinant [69] is only one of the ways in which recombinant poxviruses may become useful for vaccination.

There has already been a vast amount of molecular research on the poxviruses and poxvirus replication, none of which is covered in this review. This can be expected to continue and expand, especially as vaccinia, variola and ectromelia viruses have now been completely sequenced. It is likely that research will focus on immune evasion and the immunomodulatory genes found in all poxviruses, work for which ectromelia virus is well suited since it is a natural and virulent pathogen of that most convenient of all laboratory animals, the mouse.

In relation to the orthopoxviruses, the most worrying problem on the horizon stems from the fact that although the USSR signed the 1972 Convention on the Prohibition, Production and Stockpiling of Bacteriological (Biological) and Toxin Weapons and their Destruction (BWC), there is good evidence that research on the use of variola virus (and other agents) for nefarious purposes had been proceeding in the former USSR between about 1960 and 1992, on a very large scale [70]. This poses major political and moral problems, which it is impossible to discuss here. Nevertheless, it is important to realise variola virus may no longer be confined to the WHO Reference Laboratories, but may be available in several places in the Russian Federation and perhaps in other countries. This realisation places responsibilities on those countries or laboratories that hold such viral stocks that if they are used, the perpetrators would be committing a crime against humanity. The deferral in 1999 was specifically designed to allow countries to prepare for the use of variola virus for nefarious purposes, by setting up larger national stocks of vaccinia virus for use as a vaccine and for carrying out research with the aim of producing an effective antiviral agent.

9The European rabbit as a pest animal

All microbiologists know of the European rabbit (Oryctolagus cuniculus) as a very useful laboratory animal. However, in places like Australia and New Zealand, in which the fauna and flora have evolved for millions of years without exposure to placental animals (except a few rodents, which probably floated in on driftwood), the introduction of rabbits by European settlers in the nineteenth century was a major disaster. A few domestic rabbits were introduced from the earliest days of settlement (1788), but they failed to survive in the wild. Determined to enjoy the sport of rabbit shooting, in 1859 a wealthy grazier shipped two dozen wild rabbits from England to his property in southern Victoria [3]. They flourished ‘exceeding well’ and aided by carriage by other graziers to their own properties, they soon spread over south eastern Australia and somewhat later to all parts of subtropical Australia [71]. Within 10 years of their introduction they were recognised to be a major agricultural pest and in 1888 several of the Australian colonies and New Zealand combined to offer a very substantial reward for an effective method of rabbit control [72]. Many entries were received, including one from Louis Pasteur, but all were rejected. Rabbits continued to be a pest and a great variety of methods were used to try to control them. One of these, the most successful so far, was a poxvirus discovered in South America in 1896 [16] and it was the introduction of this virus into wild rabbit populations in Australia and Europe that explains the origins of the second part of this essay.

10The genus Leporipoxvirus

As indicated in Table 1, the genus Leporipoxvirus is one of the eight genera in the subfamily Chordopoxvirinae. The genus contains several members (Table 3) of which the most important in the present context are the two strains of myxoma virus (Brazilian and Californian) and rabbit fibroma virus.

Table 3.  Viruses of the genus Leporipoxvirus in their natural hosts and in the European rabbita
  1. aLesions in natural hosts are usually localised, benign fibromas; squirrel fibroma virus may produce generalised fibromatosis.

VirusNatural hostEndemic areaClinical signs in European rabbit
Brazilian myxoma virusSylvilagus brasiliensisSouth and Central Americageneralised, lethal disease, gross external signs
Californian myxoma virusSylvilagus bachmaniwestern USA, Baja Californiageneralised, lethal disease, often few external signs
Hare fibroma virusLepus europaeusEurope (? now extinct)localised benign fibroma
Rabbit fibroma virusSylvilagus floridanuseastern USAlocalised benign fibroma
Squirrel fibroma virusSciurus carolinensiseastern USAlocalised benign fibroma
Western grey squirrel fibroma virusSciurus griseus griseuswestern USAnot tested

11The natural history of myxoma virus

What we here call the Brazilian strain of myxoma virus was discovered by Sanarelli [16] and subsequently studied in Rio de Janeiro by Aragão [18], who showed that the natural host was the tapeti or tropical forest rabbit (Sylvilagus brasiliensis), which has an extensive range in south and central America.

Farmers in California trying to raise European rabbits on a commercial basis had long been worried by ‘mosquito disease’, which was shown to be myxomatosis in the 1930s. However, studies carried out much later [73,74], showed that although it clearly belonged to the species ‘myxoma virus’, it was different from the Brazilian strain in many characteristics. Its natural host is another species of Sylvilagus rabbit, Sylvilagus bachmani, which occurs in California, Oregon and Baja California [75].

12Myxoma virus for the biological control of rabbits

In 1918 Aragão proposed to the Australian Government that myxoma virus should be used to control the rabbit pest, but the quarantine authorities refused entry. After some discussion and consultation [3], it was allowed in, under quarantine, in 1937, and preliminary trials were carried out on its specificity and ability to be spread by mosquitoes and fleas [76]. However, it was not until 1949, when a special Wildlife Survey Section was established in the principal government research agency, the Commonwealth Scientific and Research Organisation (CSIRO), that it was given a proper trial. After earlier disappointing results, in late 1950 it spread far beyond the trial sites, producing enormous mortalities among the very numerous wild rabbits. Although there was an excellent group of zoologists and ecologists in CSIRO studying the disease in the field, nobody in Australia was studying the virus in the laboratory. I had just returned to Australia after some time overseas and decided to make the study of the virology of myxomatosis my principal research interest and appointed two young graduates who collaborated with me on the studies for the next 15 years [31].

Perusal of the literature and collaboration with an electron microscopist [77] showed that myxoma virus was a poxvirus, which made it even more attractive because of my previous experience with ectromelia virus. In a number of papers we described various aspects of the disease: its pathogenesis, using the same experimental design as with mousepox, active and passive immunity, methods of titration, in rabbits and on the chorioallantoic membrane (Burnet's method), vaccination with fibroma virus for the protection of laboratory rabbits and, with the collaboration of M.F. Day of the CSIRO Division of Entomology, mechanisms of mosquito transmission. Transmission turned out to be mechanical; the mosquito, which field workers had found to be the principal vector in Australia, was a ‘flying pin’ rather than a vector in which the virus multiplied (summarised in [2,3]).

This work was done to provide background knowledge for a study of changes in the virulence of the virus and the resistance of the rabbit in the field, a study of evolution in action. To do this our small team had splendid collaboration with the zoologists of the CSIRO Wildlife Survey Section, under the leadership of F.N. Ratcliffe. Over a period of many years the ANU group carried out regular testing of the virulence of field strains. We did this by inoculating six laboratory rabbits with a very small dose of each field strain, after a single passage in rabbit skin. Since most rabbits in such small groups died, virulence was estimated on the basis of survival times, except for tests with larger numbers of rabbits on a few selected strains, designed to confirm the validity of using survival times in groups of six rabbits as a surrogate for lethality [78]. Initially there were changes in the virulence of myxoma virus recovered from the field such that the most common strains were somewhat less virulent than the virus that had been released, which field and laboratory studies had shown had a case-fatality rate (CFR) of over 99%[79]. It was clear by 1957 that in the field the ‘99% CFR strain’ had been replaced with a less virulent ‘90% CFR strain,’ primarily because rabbits infected with the latter strain lived longer and were therefore more likely to present infectious lesions during the winter, when mosquitoes were uncommon.

The other consequence of the survival of some 10% of infected rabbits was that this left enough breeding animals to favour the selection of genetically more resistant rabbits. To test for such a change, many aliquots of the 90% CFR strain were stored in liquid nitrogen and used to challenge batches of wild rabbits that were captured in spring each year, before the beginning of the myxomatosis season. These animals were kept until they were about 16 weeks old, so that they reacted as adult rabbits and had lost any maternal antibody before being challenged. Over a period of only 7 years there was a dramatic fall in the mortality and severity of the disease in wild rabbits, although each year control laboratory rabbits died within the expected times [80]. Subsequent studies in both Australia and Europe, done by other scientists, showed that innately resistant rabbits emerged wherever myxomatosis had been present for a decade or so. Further, there was a dynamic equilibrium between viral virulence and rabbit resistance, such that more virulent viruses were more frequently found in areas where the rabbit resistance was highest [81]. Coevolution has proceeded even further during the succeeding 30 years and CSIRO scientists have shown that the majority of strains of virus currently being isolated from wild rabbits are universally and even more rapidly lethal for laboratory rabbits than the original virus introduced in 1950, although the disease they produce has only a 50% case-fatality rate in the wild rabbits found near Canberra in 1994 [3].

13Future research with leporipoxviruses

As with the less common orthopoxviruses, very little is known about leporipoxviruses other than myxoma and fibroma viruses, so there is room for a great deal of research on their natural history as well as their molecular biology. An interesting recent use of myxoma virus is similar to that described earlier for vaccinia and fowlpox viruses, i.e. as a vector for foreign genes. This is done where the requirement is to develop a vector which will spread naturally among rabbits, without affecting other animals. For example, in order to protect rabbits in commercial rabbitries against both myxomatosis and rabbit haemorrhagic disease, a vaccine strain of myxoma virus has been engineered so that it expresses the coat protein of rabbit haemorrhagic disease virus [82]. In Spain, where there are efforts to maintain high rabbit populations in certain national parks because they are needed as prey for endangered species such as lynx, attempts are being made to produce a similar recombinant, rather attenuated, field strain, in the hope that it will spread and immunise the rabbits against both myxomatosis and rabbit haemorrhagic disease. Another example of the use of recombinant myxoma virus on which a lot of effort has been expended by the Vertebrate Biocontrol Centre in the CSIRO Division of Wildlife Research in Canberra is the development of genetically engineered strains of myxoma virus which will sterilise rabbits [83].