Although previous researchers have recognized that substantial genotypic and phenotypic diversity exists among CPXV species, to date no one has objectively analyzed this diversity and put it into a taxonomic context relevant to the other members of the Genus Orthopoxvirus. The data herein depicts CPXV as a polyphyletic assemblage in which one of the major CPXV clades is more closely related to MPXV and the other is more closely allied to variola, camel and tatera pox viruses. This would mean that if the current CPXV taxonomic scheme is to remain an intact monophyletic clade then, all Orthopoxvirus isolates with the exception of Ectromelia virus and the new world species, Raccoonpox virus, Skunkpox virus and Volepox virus, would need to be assigned to the species Cowpox virus. A more reasonable solution is to rename the distinct CPXV lineages.
The fact that the traditionally defined CPXV have the largest genomes has led previous investigators to suggest that a cowpox-like virus is the ancestor of all OPV species 
. Furthermore, the deepest divergence within OPV clades is between the North American OPV species (Skunkpox, Racoonpox, and Volepox viruses) and the remainder of the genus, followed by Ectromelia virus 
. The deep node from which the New World OPXV species diverge, along with the results herein suggest that the shared CPXV phenotypic characteristics such as; large genome size, A type inclusions, flattened hemorrhagic lesions on CAM, and rabbit lesion morphology, uniting the traditionally recognized CPXV isolates, likely represent ancestral characteristics maintained in the distinct CPXV lineages but lost by other OPV species.
The smallpox vaccine is known to have originated in the United Kingdom, however the vaccine strains were most closely allied to CPXV isolates from Russia (CPXV-GRI) and from Finland (CPXV_Fin2000_Man), with the Austrian isolate (CPXV_AUS1999_867) joining just outside that sister relationship. This could be explained by a lack of sufficient UK isolates in our analyses, which would imply that multiple CPXV and/or VACV genotypes are circulating in that region that have not yet been identified. The pairwise identity between the coding regions of CPXV_UK2000_K2984 and CPXV_BR is 98.4%. These two isolates were collected 63 years apart, but are both from the southern region of England and the Bristol isolate is roughly 23 miles from the region where Jenner's vaccination work occurred. Thus the genotypes represented in these analyses should be representative of the CPXV strains available to Jenner. A more likely scenario is that most of the commercially produced vaccines (all of those included here) were not made from the original Jenner strain, but instead from isolates found in other regions of Europe. Taxonomically, the UK isolates from near the “type” locality (i.e. near Jenner's original collection site) have priority to keep the Cowpox virus designation. In future studies, the inclusion of more sequences from other circulating vaccinia-like strains would be of evolutionary, epidemiological, and taxonomic importance.
The Buffalopox virus and Horsepox virus isolates are imbedded within the Vaccinia-like clade. There are multiple hypotheses as to the origin of wild
VACV isolates 
. The two prevailing scenarios are 1) these isolates represent vaccine derived strains that were utilized during the smallpox vaccination campaign and subsequently escaped and became endemic in these regions or 2) they are native viruses. The topology displayed here favors the first scenario (). However, the long branch separating Horsepox virus from other VACV isolates suggests that more detailed analyses with additional sampling are needed to successfully resolve this discussion.
Both the historic and current literature describe “cowpox” as a disease with a single etiologic agent. Further, there has been confusion over the relationship between the causative agents of “cowpox” and “horsepox”, and the transition from what was historically called “cowpox” to the modern VACV-based vaccine strains. Much of this confusion results from the practice of naming agents after the host in which the resultant disease manifests. Using the TATV-VARV genetic distance threshold, CPXV isolates could be split into as many as five distinct monophyletic lineages. This threshold is likely conservative and is higher than the value seen between TATV and CMLV, two commonly accepted distinct OPXV species. If the lower TATV-CMLV threshold is used then Austrian, Norwegian, and British isolates are also distinct. This latter scenario would taxonomically differentiate the most geographically distinct isolates. The nucleotide data in and the coding data in support the elevation of up to four new species designations within the traditional CPXV group. According to taxonomic priority our clade 3 (&) would maintain the designation Cowpox virus
since it contains isolates from the United Kingdom closest to the type locality from which the original CPXV isolates were described. If the taxonomic changes proposed herein are correct then it is likely that yet to be discovered biological similarities exist which unite the depicted clades/species. We chose to examine three genes that have been relatively well characterized in previous studies, for evidence of amino acid level changes which unite the proposed clades and found supporting data in all three loci. The nonsynonymous mutations seen in the coding sequence data for CrmC, viroceptor, soluble virus tumor necrosis factor receptor homolog 
; soluble IFN α/β receptor homolog 
; and vaccinia growth factor, epidermal growth factor-like protein 
indicate that there are likely common protein structures shared within members of clades 1, 3 and 5 which could impact virulence and/or host range, thus supporting their recognition as distinct lineages. Furthermore, all members of clade 3 share a truncated C5L gene. We anticipate that future studies will reveal many other unique biological properties within each of the clades identified by this research effort.
In the most conservative view, our results show that CPXV strains group into at least two separate strongly supported and deeply divided clades. One clade includes VACV strains (both wild and laboratory propagated) while the other includes only strains previously identified as CPXV. This is congruent with previous observations differentiating CPXV Brighton Red and Grishak into two species 
. Additional virological characterization is needed to specifically examine and subsequently assign these clades to the correct taxonomic category, however, all scenarios described herein suggest that a re-classification of CPXV strains is clearly warranted. Historic literature states that the original vaccine strains were derived from “cowpox” samples 
. Additionally, the phylogeny presented here suggests the progenitor of the modern smallpox vaccine likely lies in mainland Europe and not the UK.