In the current study we utilized a genomic nucleotide distance-based method previously used for identifying phylogenetic clades and applied it to detect viral species. The results suggest that herpes simplex viruses isolated from Lion and Pig-tailed macaques should be designated as separate species. To our knowledge this is the first time this technique was been applied to virus species and may be useful in detecting cryptic viral species.
Host-Virus Co-speciation
Herpesviruses have been shown to cospeciate with their hosts [46], however they can cross species barriers [47], especially in captivity [48-55]. These captive transmissions, especially between macaque species can complicate phylogenetic analysis. In particular, cross-species transmission appears to be fairly common among the core herpes B strains, and has been discussed previously in depth by Eberle et al [12]. In some of the herpes B strains, the original source of the virus appears to be unclear. For instance, the cynomolgus macaque derived strain E90-136 is more distant and phylogenetically separated from the core herpes B strains (Figure 1), however it was not sufficiently distant (Figure 2) to be considered a separate species. Interestingly, strain E90-136 was isolated from a Cyno macaque which died due to a disseminated infection caused by the virus [56]. Herpes B strains are generally asymptomatic within the natural host, which may suggest that Cyno macques are not the natural reservoir for this particular viral strain. For other OWM strains, interspecies spread is well documented. The isolate 8100812 was originally isolated from a DeBrazza monkey, however restriction digest patterns showed that the Lion-tailed macaque was the natural host [51]. Phylogenetically, this appears appropriate as strain 8100812 forms a node with the two Pig-tailed macaque isolates (Figures 1A and B), and importantly matches phylogenetic profile of the macaque species themselves (Figure 1C). The correlation between Lion and Pig-tailed viruses and macaque phylogeny strongly suggests host-virus co-speciation. Additionally, while natural cross-species viral transmissions between animals does occur [47, 57-59], natural species viral transmissions between the animals and viruses in this study are fairly unlikely given the natural host ranges of the monkeys (Figure 5). The reduced likelihood of natural cross species transmission is important as it increases the probability of host-virus co-evolution. Further, for example while Lion-tailed and Bonnet macaques ranges overlap, different living strategies (frugivorous and arboreal vs generalist in human dominated environments respectively) [60, 61] between these animals make cross transmission unlikely.
Viral Species Concept
Standard definitions of what constitutes a biological species, such as a reproductively isolated population [16], are insufficient for viruses as they replicate, but do not reproduce like other organisms. Originally, viruses were simply classified according to the host that was infected, i.e. bacterial, plant or animal [62]. It wasn’t until 1950 that official principles of animal virus classification were established, with categories such as morphology, chemical composition, method of transmission, tropism and symptomatology [62]. In 1963 the International Committee on Nomenclature of Viruses (ICNV) was established and in 1966 the body proposed a taxonomic framework and classification rules which included class, order, family. This organization is now known as the International Committee for Taxonomy of Viruses (ICTV) [62, 63]. In 1990 the ICTV established an official definition of viral species which was stated as “a virus species is a polythetic class of viruses that constitutes a replicating lineage and occupies a particular ecological niche” [64], and has since evolved to state “a monophyletic group of viruses whose properties can be distinguished from those of other species by multiple criteria….not limited to natural and experimental host range, cell and tissue tropism, pathogenicity, vector specificity, antigenicity, and the degree of relatedness of their genomes or genes [65]. While this statement recommends distinguishing properties for determining species, the process is still ambiguous.
We chose to focus our efforts on genomic distance in order to apply a quantitative measure to delimit viral species. Several species delimitation methods have been used in bacteria and eukaryotes. One of the most common and recent methods for species delimitation in bacteria and eukaryotes is generalized mixed Yule coalescent, where branching patterns of a single tree transition from Yule process inter-species branching to coalescent process intra-species branching [33]. Single loci can be used for this method, however more recently multiple genes and morphological characters can be used [66]. Previously, a distance method based on gene homology and sharing was used to reevaluate viral family classifications [67]. A relatively simple genomic distance cutoff method has been used to validate viral clades [34-37] and was applied to delimit species in the current study. While the kernel density plot combined with genetic distance cutoff method described here is simplistic compared to the computation heavy generalized mixed Yule coalescent method, whole genomes and therefore more phylogenetic signal is available for analysis. We did not compare the various species delimiting methods to the genetic distance cutoff method as this was beyond the focus of the study. A caveat with the distance cutoff value used in the current study is that the cutoff value is not universal, but dataset dependent. A potential issue with using the distance cutoff method to establish species boundaries is that as the genomes of additional viruses are sequenced, the species cutoff value could potentially shift, resulting in species cutoff values that could vary over time. A general complication of the method used in the current study and in other genetic data delimitation techniques is that the methods may be delimiting populations, and not necessarily species [68]. We cannot eliminate this possibility in our analysis however this is unlikely given the large distance values between species in the dataset.
In our study to determine if the Lion and Pig-tailed derived simplex viruses were species separate from herpes B, we included all sequenced Old World monkey strains in an effort not to bias the results and establish a general cutoff for the Old World monkey group. The results of our study showed the genome-based genetic distance between Lion/Pig-tailed macaque derived viruses and the core herpes B strains were both approximately 14%, which was actually greater than the distance observed (~10%) between SA8 and PaHV-2 (Figure 2B), previously established viral species. The recovery of SA8 and PaHV-2 as separate species helps to validate the method. Both of these values were well above the species cutoff value (8.94%; Figure 2B). The genetic distance data, and the data supporting co-speciation of the Lion and Pig-tailed macaque viruses reinforces the idea that these should be designated as separate, individual species from herpes B, and each other.
Cryptic Viral Species
The term cryptic species is related to similar concepts such as sibling species, species complex, and superspecies, with the definitions between these concepts often blurred. Cryptic species are generally defined as species which appear virtually identical phenotypically, but belong to different taxa, and were thus “hidden”. It should be noted that it is not unusual for non-viral cryptic species to have some morphological differences in terms of color, size, and markings [69, 70]. Cryptic species were originally described three centuries ago [28, 29], and with modern molecular techniques have been increasingly identified across multiple organisms [71-75]. To our knowledge, the concept of cryptic species has not been applied to viruses, however species complex occasionally has [76, 77]. From the phylogenetic network of the Old World monkey simplex viruses (Figure 1A), these viruses could be described as a series of species complexes (i.e. a group closely related viruses that are difficult to separate), one comprising the macaque viruses and a second encompassing the baboon simplex viruses. The genetic distance cutoff method may be useful in establishing species boundaries in these complexes, as the method confirmed species status for the baboon derived PaHV-2 and SA8. Importantly, the method identified Lion and Pig-tailed simplex viruses as separate species (Figure 2), defining these viruses essentially cryptic species. The genetic distance cutoff method provides a quantitative threshold to determine species status and could be another tool for establishing species status among viral cryptic species complexes. Challenges and Issues
There are multiple challenges in defining species, for example recently, even in defined species, fertile hybrids among plants, birds, fish, and even mammals are not uncommon [78-81], suggesting reproductive barriers may not always separate species. This may call into question as to what constitutes a species. As previously stated, viruses do not reproduce per se, however they do recombine, and herpesviruses have been shown to be highly recombinogenic [26, 82]. Several recent studies have found natural interspecies recombinants between HSV-1 and HSV-2 [18, 20], although they share approximately 70% sequence similarity [83]. While natural recombinants between OWM viruses, which have lower genetic distances than HSV-1 and 2 have not been reported, it seems reasonable to assume it is possible. It is therefore unlikely that the ability to recombine in a host would be a factor in defining species in primate herpes simplex viruses.
Species defining methods related to virion morphology, serology, as well as gene homology and function are problematic in primate herpes simplex viruses as these characteristics are highly similar, with one of the only differences being the apparent lack of γ134.5 in the Old World monkey simplex viruses [84]. Virus morphology in particular is difficult to distinguish between simplex viruses, as an older study found that virion morphology is nearly the same between HSV and herpes B, however there may some minor differences in morphogenesis [85]. Further, to our knowledge, differences in virus morphology between in the various herpes B strains has not been investigated. From the studies performed so far, the herpes B strains examined here appear to be nearly identical in nearly every respect, including the ability to infect multiple monkey species. Future studies may be able to detect morphological difference in the viral virions or at the protein structure level. Pathogenicity is one determinative method in which there appears to be a difference between the Pig and Lion-tailed macaque viruses and the remaining herpes B strains. Studies performed by Eberle et al examining the lethal dose (LD50) of the sequenced herpes B strains in mice showed that the Pig and Lion-tailed macaque simplex viruses had different lethality phenotypes compared to the remaining herpes B strains [12]. Importantly, the LD50 values for the Pig and Lion-tailed viruses were >107 PFU, while the average for the remaining herpes B strains was approximately 104 PFU. In addition to the species delimiting method described here, pathogenicity differences support separate species designations for the Pig and Lion-tailed macaque simplex viruses.
Implications of Separate Species Designation
There are several related scientific threads derived from giving species designations to the Pig and Lion-tailed macaque species. The first is acknowledging that these viruses are on separate evolutionary paths from each other, and from other herpes B strains. This may result in closer examination of possible phenotypic differences between herpes B strains, and among other groups of closely related viruses. Further, possible future transcriptomic or proteomic data conclusions from core herpes B strains for example will not be assumed for the Pig and Lion-tailed herpes simplex viruses and would require separate experimentation.
Herpes B Core Phylogeny
Phylogenetic analysis of the remaining herpes B strains showed a core group, designated Core herpes B, containing two main clades (Figures 3A and B). Core herpes B clade 1 contained strains with longer branch lengths compared to clade 2, with strains derived from M. mulatta, M. radiata (strain M12-O), and M. fuscata (strain 7709642). It is unclear why the branch lengths are longer in clade 1, however the isolation locations and host species are variable [12] and may contribute to the greater genetic distances. The strains comprising clade 2 were all isolated from M. mulatta, and from two locations. It is possible that clade two represents a rhesus only strain grouping. herpes B core strain 9400371 may represent a sole member of a third clade, with genetic distances from clades one and two that were above the cutoff threshold (Figure 3D). The original host for this virus is unclear as the Genbank annotation (KY628983.1) states that it is rhesus macaque, however the corresponding publication [12] states it is from a cynomolgus macaque. If strain 9400371, is derived from a cynomolgus macaque, future research will help determine if is the first member of a cynomolgus macaque clade.