Duclos et al. The European rabbit likely evolved in the Iberian Peninsula and southern France from where it was distributed around Europe often as semi-domesticated stock, managed for meat, fur, and hunting, that later reverted to the wild.
The current domestic breeds were developed from these rabbits [ 41 , , , ]. Myxomatosis was suggested as a potential biological control as early as [ 8 , ] and trials were conducted during the s to test lethality and species-specificity [ 51 , 52 ]. These were followed by field releases that failed to show potential for MYXV as a biological control because of limited dissemination in the rabbit population [ 5 , 8 , ]. However, the initial field trials were conducted in very dry country where mosquitoes were absent, and further trials were subsequently conducted in higher rainfall zones.
This was facilitated by extensive inland flooding, allowing mosquito breeding in normally dry country, and the presence of huge numbers of completely susceptible rabbits at high population densities. It was also likely deliberately spread by landholders transporting infected rabbits, and there were other deliberate introductions outside the main epizootic but these seem not to have spread [ 3 , 4 , 8 ]. Natural spread of the virus in this second season was augmented by widespread inoculation campaigns and field studies were set up to monitor the outcomes of this accidental experiment.
Introductions were also made to Tasmania and Western Australia, which were geographically separated from the initial epizootic [ 8 , ]. The number of rabbits in Australia pre is not known but it was probably in the hundreds of millions and may have been an order of magnitude higher, as numbers had risen significantly in the immediate post-war period. Similarly, the reduction in population size during the first epizootic of — could not be measured, and only part of the continent was affected, but Ratcliffe et al.
They estimated that tens of millions of rabbits were killed, rather than hundreds of millions, but cautioned that they had little basis for the estimation. The virus had been maintained by periodic passage in laboratory rabbits for over 40 years; case fatality rates CFR for laboratory and wild rabbits were estimated to be as high as Once ecological studies were underway it became apparent that field strains of MYXV were emerging that were slightly attenuated compared to the progenitor SLS, and that they tended to outcompete SLS when both viruses were present [ , , ].
While SLS was not a cloned virus in the sense of having been pock-purified on the chorioallantoic membrane of fertile eggs or cloned by limit dilution in rabbits, Fenner and Marshall [ 16 ] stated that there was no evidence that it was a mixture of virus populations based on repeatability of disease produced by inoculation of very low doses of virus.
In addition, tests of clones prepared from single pocks did not produce any prolonged survival times [ 8 ]. If this is correct, then attenuated field strains must have been initially derived by random mutations from SLS. This raises an interesting question of how the attenuated viruses arose, and were initially transmitted, particularly if this trait is polygenic. Attenuated viruses must be slightly less competitive, such that they will be selected against, in the rabbit in which the mutation first arose, and mosquito transmission will generally be associated with a substantial population bottleneck because only very low numbers of viral particles are delivered [ 55 , ].
This implies that minor variants will usually be lost by chance. The virulence of field isolates of MYXV was systematically analysed by infecting small groups of laboratory rabbits, usually five or six, and at least four months old, with low doses of each virus.
Virulence was classified into five broad grades based on the CFR, average survival time AST and clinical signs [ 8 , 16 ]. The grade 3 classification was subsequently divided into 3A and 3B [ ] Table 5. This was adjusted to allow for survivors either by the method of Sampford [ ] or by allocating survivors a survival time of 60 days [ 16 , ]. Grade 1 viruses such as SLS quickly became relatively rare in the field and the predominant viruses were of grade 3 virulence with smaller numbers of grade 4 viruses Figure 1.
Where virulent viruses were isolated it was generally within a short time of release of SLS for biological control [ 16 ]. This situation persisted for the next 30 years [ ]. While there has not been any systematic examination of field strain virulence since the s, limited testing on small numbers of field strains from the s showed that highly virulent grade 1 and attenuated grade 4 and 5 viruses were still present in the wild rabbit population [ , ]. Virulence of Myxoma virus isolates in Australia — Numbers above the bars indicate the number of isolates tested for each period.
Grade 3A and 3B viruses are combined as grade 3. Data are from [ ]. Figure is reprinted from Antiviral Research [ 64 ] with permission from Elsevier. MYXV can be transmitted by direct contact; virus is shed from eroded cutaneous lesions and mucosal sites such as conjunctivae and nasal passages and can be inoculated into the upper respiratory tract or conjunctivae during social interactions or cutaneously by fighting.
However, for epizootic spread, mosquitoes or other mobile, biting arthropods are the most effective means of transmission. Mosquito transmission studies using laboratory rabbits infected with viruses of different virulence grades demonstrated that Grade 4 viruses were the most transmissible [ ] Figure 2.
Grade 5 virus was poorly transmitted probably because the rabbit immune system was able to quickly control virus replication. The grade 5 virus, neuromyxoma, used in this study may not be representative of field strains as it reaches very low titres in the epidermis; other grade 5 viruses have higher titres but are still relatively quickly controlled in the epidermis [ 16 , ]. Mosquito transmissibility of Myxoma virus. The percentage of mosquitoes that successfully transmitted virus following feeding on cutaneous lesions of rabbits infected with grade 1, grade 4 or grade 5 viruses is shown.
Despite the transmission advantage of grade 4 viruses in laboratory rabbits, grade 3 viruses predominated in the field. The most likely explanation for this is that rabbits were undergoing intense selection for resistance to myxomatosis. The emergence of attenuated viruses allowed some survival of infected rabbits, although even in the first epizootic some rabbits recovered from infection [ , ]. Over time the breeding population would be dominated by rabbits that had survived myxomatosis potentially because of allelic variants that increased resistance.
This selection may have been accelerated by high summer temperatures, which dramatically increase survival rates of rabbits infected with attenuated viruses especially rabbits with some genetic resistance [ , ]. In much of Australia, rabbits have an annual breeding season, during which does may produce 4—5 litters; the generation interval is approximately 12 months.
This was measured by infection of non-immune rabbits from each generation with the KM13 strain of MYXV under laboratory conditions [ , ]. Resistant rabbits have reduced disease severity and higher survival rates, rather than resistance to infection, and resistance can be overcome by viruses of higher virulence such as Cal MYXV or by suppressing the Th1 immune response [ 17 , , ].
The genetic basis of resistance is not known but it appears to be due to an enhanced innate immune response that controls virus replication in tissues distal to the inoculation site, effectively reducing the virulence grade of the infecting virus, and allowing development of an adaptive immune response that eventually clears the infection [ , , ].
There is also the suggestion of a temporary enhanced resistance in rabbits born to does that had mated with males that had recovered from myxomatosis even if this male was not the sire of the kittens tested [ , , ]. No experiments have been done to test these observations, which are based on post-hoc analyses of challenge data, and it is difficult to envisage a mechanism, perhaps epigenetic, by which this resistance might occur.
Although only limited studies were done, resistance does not seem to have emerged uniformly across Australia. It developed more rapidly in the hot dry areas, such as the Mallee in Victoria, than in cool moist climates [ , , ] possibly due to less frequent epizootics in the cooler areas [ 57 ].
This means that selection pressure for transmission and hence virulence may have differed among climatic zones, and there is some evidence that more virulent viruses were more prevalent in the hotter dry regions where rabbits were generally more resistant [ ], supporting the notion that transmission and virulence are strongly linked. Modeling simulations suggest one MYXV strain should predominate: grade 4 in populations with no resistance but grade 3B in more resistant populations, which generally fits the observed field data [ , ].
However, there are obviously multiple virulence grades surviving in the field. This may occur because less common strains are only excluded over a long period [ ].
Experimental releases of the virulent Lu strain [ , , , ] showed that field strains outcompeted the released virus suggesting that long-term coexistence of very virulent and moderately attenuated viruses is not likely at a local level. However, varying levels of genetic resistance and environmental patchiness may allow persistence of multiple virulence grades at some geographic scale [ ].
Alternatively, viruses may switch between virulence grades relatively readily, and the most successful will depend on the immediate local conditions and resistance.
Obviously, at a local scale, a highly lethal virus could exclude other viruses simply by being first to invade and eliminating the susceptible rabbits. An important nexus between virulence and transmissibility was further demonstrated by experiments with flea transmission of MYXV in wild rabbits.
Rabbits that survived infection, regardless of the actual virus virulence, were relatively poor sources of virus based on the proportion of fleas that were able to transmit infection from the infected rabbit. The best transmission was from those rabbits that survived for several weeks with active myxomatosis but ultimately died Figure 3.
These were the rabbits infected with grade 3A viruses; survivors infected with the same viruses were poor sources of infectivity as, in general, were the rabbits infected with grade 1, grade 3B or grade 5 viruses [ ].
Transmission of MYXV by fleas days after infection with viruses of different virulence grades. Fleas were combed from rabbits infected with viruses of virulence grade 1, 3A, 3B or 5, and tested for transmission by feeding on an uninfected rabbit. Rabbits have been grouped by survival time irrespective of the virus with which they were infected. The average percentage of infective fleas between days 8 and 28 after infection is shown for each group. Transmission is obviously a central component of virus fitness.
The drivers of arthropod transmission are the duration of high titres of virus in the epidermis, whether the infected animal survives and the survival time. Importantly, the same virus will have a different disease phenotype and transmission outcome in resistant wild rabbits compared to unselected domestic rabbits, but even within wild rabbit populations individual levels of resistance will vary.
In addition, biotic and abiotic factors that suppress the rabbit immune response, such as, age of the rabbit, nutritional stress, cold, parasites or intercurrent disease may lead to a more virulent phenotype and hence alter transmission potential [ 15 , , , , ].
Thus, under some circumstances, a highly attenuated virus could have quite successful transmission locally but be unable to invade a more resistant population [ ]. In June , the owner of an estate at Maillebois in north western France inoculated two wild rabbits with MYXV in an attempt to control rabbits on his property.
Unlike SLS, it had undergone very few passages in European rabbits since its isolation and infected rabbits had different lesion morphology to SLS with large purplish protuberant cutaneous lesions. Some dissemination was clearly deliberate, such as the introduction into Britain in [ ]. However, unlike in Australia, strenuous efforts were made to control and disrupt the spread of MYXV in Europe but to no avail [ 40 , 41 , 57 , ]. As well as spread into wild rabbits, MYXV also had significant impacts on the large rabbit farming industry, which produced domestic rabbits for meat and fur.
Rabbits in Europe carried the European rabbit flea Spilopsyllus cuniculi , which is an efficient vector of MYXV, and other fleas, with more limited geographic distribution, such as the Spanish rabbit flea Xenopsylla cunicularis , were also efficient vectors.
These fleas were not originally present in Australia but were later introduced to aid in the dissemination of MYXV [ , , ]. Flea transmission of myxomatosis can occur throughout the year whereas mosquitoes and other flying vectors such as culicoides midges and simuliids black files are more seasonal. All of these vectors were important in different parts of Europe, although fleas were the main vector in Britain [ 41 , ].
Critically, while the environment, climate, vectors and progenitor virus strain were all somewhat different to Australia, the evolutionary outcome was remarkably similar with emergence of attenuated viruses and selection for genetically resistant rabbits. However, there were differences in the rates at which attenuated viruses became dominant in the field and the emergence of resistance.
In France, the first attenuated virus was not isolated until April , nearly three years after the initial introduction, whereas in Britain a grade 3 virus was isolated about 12 months after the initial outbreak [ 16 ].
Whether the difference in timing reflects lack of sampling, as occurred during the first epizootic in Australia, or different selection pressures is not clear. The intensive studies on virulence carried out by Fenner and co-workers in Australia were not replicated in Europe, although limited data from France shows that in the majority of viruses tested were of grade 3 virulence and in the majority were of grade 4 virulence [ ].
This may reflect the slow emergence of resistance in wild rabbits in France see below , or be biased by sampling from farmed rabbits that were not under selection for resistance [ 41 ].
In testing of small numbers of isolates from to , grade 3 viruses again predominated [ 40 ]. More recent testing of 20 viruses, isolated from wild rabbits in Spain between and , showed that the majority were of high virulence equivalent to grade 1 or 2 , although the testing protocol was different to previous studies [ ]. However, unlike in Australia, grade 2 viruses were always a substantial proportion of the isolates and grade 4 viruses were relatively rare especially after It was initially believed that the slower emergence of attenuated viruses was because fleas, the main vector in Britain, only left rabbits when the animal died and that this would lead to selection for more virulent viruses.
However, it was subsequently shown that fleas were highly mobile and moved freely between live rabbits [ ]. Moreover, it is clear that in parts of France and Spain mosquitoes and other flying insects were available as vectors [ 16 ]. Virulence of Myxoma virus isolates in the United Kingdom. Numbers above the bars indicate the number of isolates tested. Data are from [ , ]. In France, resistance was quite slow to emerge in the wild rabbit population, although the systematic studies done in Australia were not replicated, no resistance was apparent in [ 8 ].
As in Australia, resistance appears to have been selected to different degrees in different regions, being reported as particularly strong in the Mediterranean south [ 40 , 41 , 57 ].
Similarly, in Britain, resistance was relatively slow to emerge; testing in showed that wild rabbits challenged with a grade 3 virus had longer survival times and a slightly higher survival rate compared to laboratory rabbits but the differences were relatively minor. While resistance in Britain was widespread, the degree of resistance appeared to vary between geographic regions, as also occurred in France and Australia [ ].
The emergence of strong resistance, possibly coupled with the dynamics of flea transmission, may be responsible for the increasing prevalence of grade 2 viruses in Britain. Unlike in Australia, there was significant domestic rabbit breeding in Europe and the interchange of viruses between the wild and farmed populations may have had an impact on evolutionary dynamics of MYXV, especially when widespread but possibly inefficient vaccination was applied in farmed rabbits [ ].
Arthur and Louzis [ 40 ] consider little interaction occurs between closed modern production units and wild rabbits but this would not have been the case for the large numbers of small holdings and backyard producers, especially during the early epizootics, and also for the producers of wild and cross-bred rabbits reared for release [ 41 , ].
In domestic rabbit operations, destruction of infected rabbits to prevent transmission of myxomatosis would prevent selection for resistance [ 8 ], however, it could also inadvertently have selected for atypical clinical signs that were confused with bacterial respiratory infections and permitting ongoing transmission. However, the overall appearance of the affected rabbits can be very similar to those with the dermatrophic form, with swollen, closed eyelids, swollen heads and ears, mucopurulent rhinitis and blepharoconjunctivitis.
Direct contact is needed for transmission as rabbits housed in adjacent cages did not transmit [ ]. This amyxomatous form of disease has also been reported in wild rabbits [ 40 , ]. It should, however, be recognized that myxomatosis has always been readily transmitted by direct contact [ 51 , 52 ].
Serological and virological evidence [ , ] indicates that many cases of myxomatosis in commercial rabbitries go unrecognized, implying that there may be ongoing transmission of attenuated viruses.
Interestingly, some amyxomatous strains which caused severe disease in conventional farmed rabbits caused very mild clinical signs in laboratory rabbits maintained free of the common bacterial pathogens [ , ].
These observations suggest that MYXV strains that mimic, and may be dependent on, bacterial upper respiratory disease could be evolving in farmed rabbits. It is likely that these unrecognized cases of myxomatosis are responsible for much of the speculation about how outbreaks of myxomatosis occur in apparently isolated populations. Reactivation and shedding of MYXV in recovered rabbits has been suggested to explain some outbreaks [ ] and purportedly demonstrated [ ].
However, subsequent attempts to demonstrate reactivation of virus were unsuccessful unpublished data cited in [ 57 ] [ ]. However, incubation periods of up to 20 days for some amyxomatous strains mean that rabbits could have been incubating the disease prior to vaccination [ ]. In addition, vaccinated rabbits can become infected and shed MYXV following challenge [ ]. Evolution of both Australian and European isolates of MYXV to produce a flatter less protuberant skin lesion has been described [ 8 , , ], although these lesions are still quite prominent compared to more recent Australian isolates.
Some highly virulent modern field isolates of MYXV from Australian also produce a largely amyxomatous phenotype in laboratory rabbits with secondary cutaneous lesions being quite rare and the primary lesion at the site of inoculation extremely small and undifferentiated from the surrounding skin. However, these viruses tend to cause fairly typical myxomatosis in resistant wild rabbits [ ]. Live virus vaccines have been extensively used in Europe to protect farmed and wild rabbits from myxomatosis [ 40 , 57 , , , , ].
Initially, only the heterologous RFV was available [ 14 ], but subsequently, a variant of the Cal MYXV MSD strain, attenuated by passage in cell culture, was used [ ], and then local Lu-derived strains that were attenuated by cell passage were developed, such as the French SG33 [ ]. Vaccination has led to some interesting evolutionary outcomes, for instance, the MSD vaccine strain appears to have established a transmission cycle in farmed rabbits [ ]. In addition, the SG33 vaccine strain has undergone multiple recombination events with a Cal MYXV, presumably the vaccine strain [ , ].
This is similar to rates estimated for variola virus, the orthopoxvirus that caused smallpox in humans [ ]. Phylogenetics and phylogeography of Myxoma virus in Australia and Europe.
Previously determined virulence grades are indicated as 1 to 5. Bootstrap values are shown for key nodes, and all horizontal branches are drawn according to the number of nucleotide substitutions per year.
Credible intervals for divergence times for two key nodes are also shown see [ , ] for details. It has 72 nt differences from Lu including indels but only counting one TIR [ ]. Whether these mutations were present in the original isolate or arose during rabbit passage cannot be determined. As already noted, while both SLS and Lu are of grade 1 virulence when tested in laboratory rabbits, Lu is actually more virulent than SLS when tested in resistant wild rabbits [ ].
In addition, the two viruses have substantially different lesion morphology with Lu inducing large purple cutaneous tumours whereas SLS tends to induce a flatter although still raised skin lesion. Other South American isolates with limited rabbit passage history tend towards the Lu lesion morphology and the virus strain from which SLS was derived also tended towards a purple type lesion [ 16 ] suggesting that some mutations occurred during its passage history prior to release in Australia.
The M protein has a significant role in virulence by downregulating MHC-I from the surface of infected cells [ , , ]. The function of M is not known but a gene knock-out virus was attenuated [ 91 ], although, Glenfield is more virulent than the progenitor SLS [ ]. Although only six isolates were tested, by the end of the second epizootic in the autumn of April and May , the viruses from a wide geographic area included slightly attenuated grade 2 and more attenuated grade 3 viruses but only two grade 1 viruses, although one of these showed prolonged survival in one rabbit [ 16 ].
Unfortunately, none of these viruses were available for sequencing. Ur was originally described as the prototype Australian grade 4 strain [ 16 ], but in all recent testing has been grade 5 virulence [ , ]. All three early isolates have the three indels found in SLS, confirming that these were definitely present in the progenitor virus; Ur and KM13 also share the indel in ML found in Glenfield, suggesting that this mutation was relatively widespread.
Both viruses have only low numbers of other substitutions but interestingly four amino acid substitutions in KM13 are present in all recent Australian isolates sequenced [ ]. There are no Australian sequences for the period between and However, genome sequences for 21 viruses from to have been completed. Virulence classifications are known for six of these, with three viruses classified as grade 1 virulence and three of grade 4 or 5 virulence [ ].
Molecular clock studies suggest that these 21 isolates share a common ancestor that was present between and [ ]. Despite the high mutation rate, much of the genome is highly conserved: across all the Australian sequences 48 genes have no change and a further 23 have only synonymous substitutions. This small number of mutations in individual genes means any analysis of gene and site-specific selection pressure is highly problematic, in turn making it very difficult to determine which mutations have been responsible for initial attenuation and potentially reversion to virulence.
There is no evidence for recombination in these data. A common trend is the disruption of ORFs by single or multiple nucleotide indels, which frequently occurs in homopolymer sequences; in the Australian isolates 13 of 16 single nucleotide indels within coding sequences occur in homopolymer tracts Table 6. Both the ML indel and the ML mutations occur on the long branch of the tree separating the — viruses from the viruses.
Some virus isolates have further disruptions to the ML ORF, or mutations forming stop codons, suggesting that it is becoming a pseudogene Table 6. Whether this duplication of two potential immunomodulatory genes has an effect on viral fitness or virulence is not known.
Fragmentation or duplication of genes has the potential to allow evolution of new functions, for example, the CPXV12 gene in cowpox virus, is a gene fragment which has evolved a novel role in MHC-I downregulation and hence avoidance of T cell recognition [ ]. The first reported outbreak of myxomatosis in Britain was in October ; sequencing of three viruses from the early radiation: Cornwall, April grade 1 ; Sussex, September grade 3 ; Nottingham, April grade 5 , showed that the virulent Cornwall had eight nucleotide differences from the Lu progenitor, seven of which were in ORFs and five were non-synonymous [ , ].
However, assuming that the British viruses were derived from a single introduction, divergent lineages had already arisen by , since most of the mutations in Cornwall are not shared by the other two early viruses sequenced, and Sussex has mutations that are not present in Nottingham. It seems likely that the multiple indels are responsible for attenuation of Nottingham, but whether the disruption to ML is responsible for the attenuation of Sussex is not clear as a virulent Australian virus also has this gene disrupted [ ].
Sussex shows that mutations associated with attenuation were present in the UK within less than 12 months of the initial outbreak, so that a lack of mutations was not responsible for the slower appearance of attenuated viruses in the UK. Hence these data tentatively suggest that there were different selection pressures operating in the UK compared to Australia. The only other complete genome is that of the highly attenuated Spanish strain of MYXV isolated in [ 60 ].
Four genes in are disrupted by indels: ML , ML , MR and MR , the latter two have demonstrated roles in virulence [ 95 , ], providing possible explanations for the extreme attenuation.
The indel in ML is independent of the mutation in Nottingham and the indel in ML is different to that in the Australian isolates. Outside of these indels there were 67 nucleotide substitutions between the grade 5 strain and the Lu progenitor over the 43 years of evolution in European rabbits. Viruses with the same mutation in the MR gene as have been isolated in the Netherlands and so may be widely distributed in Europe [ ] but, in Australian viruses, no mutations have been found in this gene [ ].
Interestingly, RFV has lost the M gene orthologue [ 61 ]. The poxvirus early, intermediate and late promoter sequences appear to be conserved across the leporipoxviruses sequenced [ 61 ]. However, the impact of level of gene expression on fitness and virulence is not clear. While selection for alteration in gene expression is an attractive concept for modifying virulence, the fact that in many cases promoter sequences are embedded within the 3' end of the preceding gene may tend to conserve these sequences.
Examination of potential promoter sequences across all of the MYXV genomes sequenced revealed only six mutations and most of these would be predicted to only have minimal effects based on previous mutational analysis of the vaccinia virus promoter sequences [ , , ]. Overall, the conclusion from the genome studies is that in these large, complex DNA viruses there is no common pathway to attenuation or virulence, but a convergence for phenotype that is compatible with multiple different genotypes and possibly involving complex epistatic interactions.
Variants would undergo local selection but, as already noted, it is likely that a broad range of genotypes is compatible with successful spread depending on the local ecological conditions and rabbit population dynamics, including the level of resistance in the population. However, it is clear that from very early in the adaptation of MYXV to European rabbits, these local variants outcompeted virulent viruses released into the same populations [ 4 , , ].
The dynamics of selection must be altered in resistant rabbit populations potentially driving selection for more virulent viruses, when measured in laboratory rabbits. The sequence data for RFV and Cal MYXV suggest that in some evolutionary pathways loss of genes is compatible with successful transmission, but whether these genes were actively selected against to enhance transmission by reducing virulence cannot be determined. Similar reductive evolution potentially associated with adaptation to new environments or selection for attenuation has been described in the orthopoxviruses [ , ].
The relatively large number of mutations seen in 50 years of evolution of MYXV in European rabbits potentially provides many pathways by which changes in virulence may occur and, as has occurred in MSW and RFV, it may be that a number of mutations are compatible with increased fitness and that which ones get fixed in the population in the new host species may in part reflect stochastic factors such as local extinction of host populations as much as active selection.
Geographically, multiple lineages of MYXV can coexist over quite small distances with particular variants dominant on sites in different years and multiple variants isolated on single study sites during a season [ , , ] Figure 5.
At face value, this suggests that there is little difference in overall fitness of these viruses at a local scale, but that regionally successful variants arise and are lost depending on the geographic scale under consideration and rabbit population dynamics [ ]. However, phylogenetically closely related viruses have been isolated over km apart in the dry hot region of SW Queensland and in the cooler, higher rainfall, Canberra district suggesting that widespread dispersion of viruses can occur.
Although the mechanism of spread over such distances is unclear there is some evidence for long distance dispersal via mosquitoes, for example, to isolated islands [ 3 , 8 ].
Following its spread in Europe, the Lu strain was released experimentally in Australia in [ ] but was outcompeted by local field strains. Despite this early failure, Lu was widely released for rabbit control in Australia from the s to the s in conjunction with the establishment of the European rabbit flea as an additional vector.
However, there is no evidence from either genomic sequencing or from restriction fragment length polymorphism analysis of hundreds of samples that Lu has established in Australia and contributed to the existing field strains [ , , , , ].
Phylogenetically, all the recently sampled viruses are distinct from SLS, Glenfield and Lu suggesting that the ongoing releases of these viruses made no significant contribution to the subsequent evolution of myxoma virus in Australia.
Sequence studies provide the basis for understanding the molecular changes that led to attenuation, although it is difficult to determine the roles of individual mutations and gene disruptions. Similarly, it is difficult to know whether virulence was maintained by circulating virulent viruses or re-evolved in attenuated viruses.
That is, do particular virulence lineages maintain stable circulation or is it relatively easy to flip back and forth between virulent and attenuated viruses with local selection determining which phenotype is successful? Further important questions include whether there is ongoing selection pressure on the virus driven by increasing resistance in the rabbit population, what capacity the rabbit has for further evolution, and what the cost of resistance is to the rabbit? This is further complicated by ongoing selection pressure on the European rabbit imposed by the virulent rabbit haemorrhagic disease virus RHDV , a calicivirus, first described in China in , that has spread in Europe and, more recently, Australia causing high mortality in rabbit populations [ , ].
Writing 50 years ago, Fenner and Ratcliffe [ 8 ] suggested two possible evolutionary trajectories. Such an outcome is dependent on the rabbit being able to evolve very strong resistance that prevents dissemination and generalized disease, limiting virus replication to the inoculation site.
This would require the virus to maintain sufficient immune suppression capacity to prevent clearance at the inoculation site but not overwhelm the rabbit. Obviously, the previously characterized grade 5 viruses would not be selected as they would be too readily controlled by the rabbit immune system. Viruses selected in these conditions of high resistance could be of very high lethality for non-resistant rabbits. The alternative is that the virus mutates to lose the ability to disseminate within the host while retaining its resistance to immune clearance from the inoculation site perhaps by recombination with RFV.
However, it is questionable how well such a virus would compete with field strains that cause generalized disease, because multiple virus-rich tissues would provide more opportunities for insect transmission than a single fibroma unless it had very prolonged persistence.
The second possibility was that selection for genetic resistance in the rabbit population reaches a plateau because further selection for virus virulence is not possible. In this scenario, a relatively stable situation occurs whereby virus strains with sufficient virulence to cause the generalized lethal disease optimal for transmission continue to predominate in the field, causing appreciable mortality in the population, but not driving further selection.
Rabbit resistance may also plateau because the costs of resistance are too high or there is insufficient genetic polymorphism for further selection in the absence of mutation in the rabbit genome. Depending on the eventual level of resistance in the rabbit population, this could see the emergence of viruses with hypervirulent phenotypes in unselected rabbits.
A third option not considered by Fenner and Ratcliffe [ 8 ] is that subtypes of the virus could evolve towards contact transmission as may be occurring in Europe with the evolution of amyxomatous strains of virus. Such transmission is likely only to be sustainable where high rabbit densities and close contact prevail.
After 65 years of evolution, there is very little evidence for the emergence of a fibroma type disease and some evidence that grade 1 and 2 viruses are more common in the field in Europe and Australia, suggesting that genetic resistance may still be driving virus evolution. Clearly, further systematic studies of viruses and rabbits are needed to understand how this complex evolutionary story will play out and to understand the biological and molecular mechanisms of this evolution.
A single model may not prevail with different environments favouring different evolutionary solutions. National Center for Biotechnology Information , U.
Journal List Viruses v. Published online Mar 6. Peter J. Read , 2 and Edward C. Holmes 4. Find articles by Isabella Cattadori. Andrew F. Find articles by Andrew F. Edward C. Elliot J. Author information Article notes Copyright and License information Disclaimer. Received Dec 31; Accepted Feb This article has been cited by other articles in PMC. Abstract Myxoma virus MYXV is the type species of the Leporipoxviruses , a genus of Chordopoxvirinae , double stranded DNA viruses, whose members infect leporids and squirrels, inducing cutaneous fibromas from which virus is mechanically transmitted by biting arthropods.
Keywords: myxoma virus, leporipoxvirus, poxvirus, myxomatosis, rabbit, coevolution. Introduction The introduction of myxoma virus MYXV into the European rabbit Oryctolagus cuniculus population of Australia is the classic example of host-pathogen coevolution following a species jump. Table 1 The leporipoxviruses. Open in a separate window. Rabbit Fibroma Virus RFV RFV, also known as Shope fibroma virus, is the best studied of the leporipoxviruses in its natural host and provides a model for their biology.
Hare Fibroma Virus Hare fibroma virus is the only leporipoxvirus naturally found outside the Americas. Squirrel Fibroma Virus Squirrel fibroma virus induces cutaneous fibromas and proliferative epidermal lesions in eastern gray squirrels Sciurus carolinensis in North America. Species Specificity of Leporipoxviruses Divergence from a putative common ancestral leporipoxvirus likely occurred either by shifts into new species and subsequent adaptation or ongoing host speciation and geographic separation.
Table 2 Genes in myxoma virus encoding demonstrated or potential immunomodulatory or host-range proteins. Gene Protein Function no. Myxoma Virus Evolution in Australia 4. The European Rabbit and Initial Trials with MYXV The European rabbit likely evolved in the Iberian Peninsula and southern France from where it was distributed around Europe often as semi-domesticated stock, managed for meat, fur, and hunting, that later reverted to the wild.
Figure 1. Figure 2. Selection for Resistance in the Rabbit Population Despite the transmission advantage of grade 4 viruses in laboratory rabbits, grade 3 viruses predominated in the field. Virulence and Transmission Modeling simulations suggest one MYXV strain should predominate: grade 4 in populations with no resistance but grade 3B in more resistant populations, which generally fits the observed field data [ , ].
Figure 3. Introduction and Spread In June , the owner of an estate at Maillebois in north western France inoculated two wild rabbits with MYXV in an attempt to control rabbits on his property. Attenuation in Europe In France, the first attenuated virus was not isolated until April , nearly three years after the initial introduction, whereas in Britain a grade 3 virus was isolated about 12 months after the initial outbreak [ 16 ].
Figure 4. Selection for Resistance in Europe In France, resistance was quite slow to emerge in the wild rabbit population, although the systematic studies done in Australia were not replicated, no resistance was apparent in [ 8 ]. Evolution of the Phenotype of Myxomatosis in Europe Unlike in Australia, there was significant domestic rabbit breeding in Europe and the interchange of viruses between the wild and farmed populations may have had an impact on evolutionary dynamics of MYXV, especially when widespread but possibly inefficient vaccination was applied in farmed rabbits [ ].
Evolution of Vaccine Strains Live virus vaccines have been extensively used in Europe to protect farmed and wild rabbits from myxomatosis [ 40 , 57 , , , , ]. Figure 5. Phylogeography Geographically, multiple lineages of MYXV can coexist over quite small distances with particular variants dominant on sites in different years and multiple variants isolated on single study sites during a season [ , , ] Figure 5. Future Evolution of Host and Pathogen Sequence studies provide the basis for understanding the molecular changes that led to attenuation, although it is difficult to determine the roles of individual mutations and gene disruptions.
Author Contributions All authors contributed to concept discussions and editing; PJK wrote the paper. Outdoor rabbits are generally at a higher risk of infection than indoor rabbits. This disease is caused by the myxoma virus, a strain of leporipoxvirus. Outbreaks of it are more more likely when mosquitoes are numerous, in the summer and fall. Your veterinarian will perform a thorough physical exam on your rabbit, taking into account the background history of symptoms and possible incidents that might have led to this condition.
A blood profile will be conducted, including a chemical blood profile, a complete blood count, and a urinalysis. One of the obvious symptoms that will help your doctor to make a diagnosis will be the presence of nodules on the skin surface.
However, in cases that are very sudden peracute , there may be no lesions. Subcutaneous ecchymoses, or purple, bruise-like spots on the skin due to the rupturing of blood vessels, are sometimes associated with myxoma virus. An internal exploration may find ecchymoses in serosal surfaces lining of the gastrointestinal tract as well. In many cases, there is hepatic necrosis death of the liver tissue , splenomegaly enlargement of the spleen , infarcts death of tissue due to deprivation of blood supply , or hemorrhage in the lungs, trachea windpipe , and thymus gland near the base of the neck.
Other findings include undifferentiated mesenchymal cells the undetermined cells that are capable of transforming into many of the materials needed by the body e. If the rabbit is pregnant when it becomes infected, necrotizing lesions may be seen in fetal placentas. Due to the serious nature of this virus, most rabbits do not survive.
Treatment is instead focused on making your rabbit as comfortable as possible. Cutaneous lesions predominate. Pathology: The most prominent gross lesions are skin tumors and pronounced cutaneous and subcutaneous edema. There is epidermal hyperplasia with intracytoplasmic eosinophilic inclusions in basal cells. Epidermal degeneration with intense inflammatory infiltrate may be seen.
The dermis is infiltrated with large stellate myxoma cells in a mucinous matrix with inflammation and neovascularization. Myxoma cells often contain intracytoplasmic eosinophilic inclusions arrowheads.
Dermal necrosis or hemorrhage from vascular compromise may be seen. Visceral lesions are occasionally observed, especially in young rabbits.
0コメント