family are large double-stranded DNA viruses with a complex structure and a broad linear genome ranging from 128 to 365 kbp. The evolutionary origin of poxviruses is still ill-defined, however, it is believed that their genomes have evolved over thousands of years through both gene gain and loss, mainly through horizontal gene transfer and gene duplication events [1
]. The saltwater crocodilepox virus (SwCRV) belongs to the genus Crocodylidpoxvirus
, a member of the subfamily Chordopoxvirinae
in the family Poxviridae
and is a known causative agent of poxviral lesions on Australian saltwater crocodile skin (Crocodylus porosus
]. Although the International Committee on Taxonomy of Viruses formally recognised crocodile poxviruses under the genus Crocodylidpoxvirus
], at this stage, no taxonomic classification has been granted for SwCRV. A recent study suggested that there were likely two major SwCRV subtypes naturally circulating in the saltwater crocodile population [3
] on one farm. However, relatively little is known about the origins, infection dynamics, genetic diversity and inter-farm genetic variability among the circulated SwCRV in saltwater crocodiles. In agreement with other crocodilian poxviruses [6
], saltwater crocodilepox viruses are morphologically similar to orthopoxvirus virions, demonstrating a brick-like shape with rounded corners and a dumbbell-shaped central core and lateral bodies [3
]. Furthermore, intracellular mature virions of SwCRV display the regular crisscross surface structure pattern, which is characteristic of parapoxvirus virions [3
Poxvirus lesions have been reported in a number of different crocodilians, including Caiman crocodilus fuscus, Caiman crocodilus yacare, C. porosus, Crocodylus johnstoni
and Crocodylus niloticus
]. In C. porosus
, poxvirus infection has been reported as a significant skin pathogen because if an individual is harvested with one or more poxvirus lesions, the lesion will result in an obvious defect on the finished leather product (4, 10). Moore et al. (4) described four stages of the poxvirus lesion development on C. porosus
belly skins. The “early active” stage is characterised by lesions that are, on average 0.85 ± 0.29 mm2
grey-white foci with normal to pin-point keratin damage. As the lesion progresses into the “active” stage, there is an obvious enlargement of the lesion as the central plug (keratinocytes containing virus inclusion bodies) increases, compacting the underlying dermis and dislocating the overlying keratin. This enlargement continues until the central plug is expelled (“expulsion” stage) into the environment and the “healing” stage begins. Throughout these stages, histology reveals that poxvirus lesions do not breach the basement membrane layer of saltwater crocodile epidermis and given enough time they will heal without detriment to the quality of the finished leather product (4, 10). However, waiting for lesions to heal extends the production time of crocodiles, and, therefore, production costs, notwithstanding the risk of more lesions developing in the meantime (10). As such, poxvirus poses a substantial financial risk to Australian crocodile producers [4
Even though SwCRV is an important pathogen of C. porosus, data regarding its evolutionary history, genetic diversity and molecular epidemiology are not sufficient due to the limited collection of only two complete genomes of SwCRV thus far. Therefore, this study was designed to firstly understand poxvirus infection dynamics, followed by developing a comprehensive sequence profile of a set of representative SwCRV genomes to identify the likely evolutional history, genetic diversity and inter-farm genetic recombination patterns across five different crocodile farms located in Northern Australia. In this study, 14 complete SwCRV genomes (12 SwCRV1 and 2 SwCRV2) were sequenced, assembled and annotated. Combined with the previous two SwCRV (3), these genomes represent a robust tool for studying the evolutionary history and genetic diversity of SwCRV and for identifying likely recombination events within SwCRV. This dataset may also offer valuable insights into the evolution of poxviruses as it represents sequence analysis of a group of highly related poxviruses in a unique environment where infection is constant and reoccurring.
Poxvirus infection in C. porosus
was first reported in 1992 [32
] and remained a significant economic risk due to increased production times waiting for lesions to completely heal thus they cannot be seen on the finished tanned leather product. Recent studies have confirmed the presence of typical poxvirus structures in the pathological lesions using transmission electron microscopy and the sequenced genome of saltwater crocodilepox virus [3
]. In this study, it was demonstrated that significantly different infection dynamics and pathogenic outcomes exist between two Australian crocodile farms and these observations were further defined by genomic sequencing, whereby distinctly different clades of functional genes were formed. To achieve this, 14 complete SwCRV genomes were constructed from lesions sourced from five different Australian crocodile. Using these, in addition to two previously reported SwCRV genomes (3), we established a well-supported evolutionary relationship among poxvirus sequences under the genus Crocodylidpoxvirus
This study determined that the grower phase of crocodile production presents the highest risk for the development of poxvirus lesions. Although these lesions will heal [4
] and not affect the skin quality, their presence delays harvest and increases the costs of production combined with the risk of more lesions developing in the meantime. Between the two farms observed, there were definite differences in poxvirus infection dynamics. Farm 2 had significantly more lesions than Farm 1 (Figure 2
and Figure 3
) and the Farm 2 crocodiles were observed to have more poxvirus lesions, in both the early active, active and expulsion stages, as they approached finishing size, conceivably deferring their movement into the higher cost/unit finishing stage of production (Figure 4
). Comparatively on Farm 1, the risk of new poxvirus lesions developing significantly decreased when crocodiles were >135cm in total length (Figure 4
). This information must, however, be considered in the context of farmed crocodiles, which are more densely located than animals in the wild, and it is possible that this factor may contribute to poxvirus infection dynamics, and that alternate pathogenesis may be seen in wild crocodiles. The differences in infection dynamics in these two farms, which display similar husbandry practices and share wild-harvested eggs from the same collection areas, nevertheless, led us to examine the genomic variation between SwCRV genome sequences on these and other farms in Northern Australia.
We have previously demonstrated that SwCRV on Australian crocodile farms is distinct from other chordopoxviruses, and thus its reservoir is unknown [3
]. Nonetheless, SwCRV has demonstrated a close relationship with Nile crocodilepox virus isolated from a different continent with no species distribution overlap between C. niloticus
and C. porosus
. Genomic analysis of 16 SwCRV isolates from five farms in Australia revealed a separation of these isolates into three distinct clades, supported by both the construction of a ML tree and a NeighborNet tree. Although there was a propensity for particular farm SwCRV isolates to cluster within clades, others such as those from Farm 2 distributed across the three different clades (Figure 6
), perhaps highlighting the likely mode of inter-farm viral transmission. In some cases, the distribution of clades may be reflected by animal transfer arrangements between farms. For example, Farm 5 supplies crocodiles to Farm 3 and these isolates clustered together within Clade-I, albeit at a different subclade. It is unknown if the isolates sampled from Farm 3 were from crocodiles originally sourced from Farm 5 or not. Further, Farm 3 supplies Farm 4. The isolates from Farm 4 clustered with SwCRV-2 (Clade-III), but it was again unknown if this isolate was from a Farm 4-raised crocodile or not. Farms 3 and 4 also shared wild-harvest egg collection areas similar to Farms 1 and 2. The distinct difference in the pathogenesis of poxvirus lesions (Figure 4
), combined with the distribution of SwCRV isolates from these farms in alternate clades, is perhaps indicative that the dominant environmental source of the poxvirus is now mostly farm-based and acquired following hatching of the crocodiles. However, further studies are needed to clarify the initial reservoir and host range of SwCRV beyond crocodile farming. Additionally, we must also keep in mind that the farmed crocodiles are more densely located than animals in the wild and that, therefore, their infection dynamics may be altered in this setting, due to the nature of the crocodile hunting and interactions with each other in the wild that this species would still be expected to undergo.
Although we observed a high degree of sequence similarity, ranging from 97.1% to 99.9% (Figure S1
), and an intact set of conversed core genes amongst all SwCRV sequences, we did observe a distinct variation in the gene encoding the IMV A type inclusion-like protein P4c (SwCRV1-188) amongst the separate isolates.
The IMVA P4c protein was fragmented due to multiple insertions/deletions in one SwCRV isolate from Farms 1, 4 and 5, and four isolates from Farm 2 (Figure S1
). At the gene level, this variation was dominated in the SwCRV genomes isolated from Farm 2, being present in 4 out of 5 isolates. The poxvirus P4c protein is a structural protein present on the surface of the intracellular mature virus particle (IMV) and has been demonstrated to be necessary for directing IMV into A-type inclusions (ATI), formed by the A_type inclusion protein (Atip) [33
]. Many orthopoxviruses embed virus particles into dense bodies, called ATIs, and it is believed that this may provide environmental protection for the virion. While many notable poxviruses, including monkeypox and variola virus, contain disrupted versions of the P4c protein (or its homolog), the lack of inclusion body formation may suggest a positive infection advantage. Interestingly, recent evidence has suggested that interruption of the cowpox P4c protein enhances the pathogenicity in the lungs of mice, as well as viral replication [35
]. Given that the P4c gene is fragmented in the majority of the Farm 2 SwCRV viral isolates, a farm where we see increased presentation of poxvirus lesions, as well as enhanced pathogenicity and prolonged infection, one hypothesis might be that this gene interruption is driving this phenomenon. However, in the absence of a tissue culture system for this virus, further experimentation will be required to assess multiple early active crocodile poxvirus lesions for both their P4c sequence information and the presence of inclusion bodies in their lesions, as well as following the pathogenesis of initial infection to determine its outcome.
Viral recombination can have a major impact on the emergence of new viruses and the expansion of viral host ranges, as well as increases in virulence and pathogen diversity [36
]. It has been well documented that recombination plays a pervasive process of generating diversity in a wide range of RNA viruses, as well as in many DNA viruses [37
]. The role of recombination in the case of viruses belonging to the genus Crocodylidpoxvirus
is still not understood due to the lack of sufficient sequence data. Interestingly, using the 16 genomes generated in this study, SwCRV genomes appear to be the subject of multiple recombination events. There was a large number (n
= 24) of potential inter-farm and/or inter-subtype likely recombination events detected among the SwCRV genomes isolated from Australian C. porosus
. Similarly, MCV, which is distantly evolutionarily related to members of the genus Crocodylidpoxvirus
, has also revealed the existence of large-scale recombination events between two different MCV subtypes [40
]. It is quite possible that the role and importance of recombination as a mechanism for SwCRV evolution may have maintained a similar pathway to MCV. The availability of more sequence data, especially from wild crocodile SwCRV lesions if found, will allow more accurate determination of these evolutionary relationships to facilitate a better understanding of the diversity observed and the variability of certain biological traits such as host range and transmissibility. These are essential factors that will influence effective management and control of this economically significant virus infection for the Australian crocodile industry.
The evolutionary origins of the Poxviridae family remain unknown although they are a very diverse DNA viral family that is exclusively cytoplasmic replicating and able to infect reptiles, humans, birds, mammals, insects and fish. One of the newer described members of the Chordopoxvirinae subfamily, within the poxvirus family, is the saltwater crocodilepox virus (SwCRV), belonging to the new genus Crocodylidpoxvirus. The saltwater crocodile is an ancient species, having evolved from the archosauria clade that includes the dinosaurs, and further insight into the evolution of poxvirus infection in these animals may offer valuable insights into evolution of the Poxviridae viral family. Additionally, the wide-spread incidence of poxvirus in farmed crocodiles may also afford the opportunity to obtain further valuable insights into natural viral selection processes in an in vivo setting.