MD is currently well-controlled by vaccination, although it previously caused serious economic losses to the poultry industry. However, the virulence and genetic characteristics of MDV strains have changed over time, and divergence seems to be correlated with the introduction of vaccines [23, 47]. According to a previous report, the genome sequences of MVD strains are classified into two main clusters, Eurasia and North America [47]. The genome sequences of MDV strains isolated in USA, China, and Europe have been investigated [33–47]. However, data on the whole genome sequences of MDV strains from other regions are limited. We previously reported changes in the genetic characteristics of meq genes in Japanese isolates [22]. In the present study, to compare the genetic characteristics of the whole genome of MDV strains circulating in Japan, we analyzed the whole genome sequence of a Japanese field strain, Kgs-c1, isolated in 2014.
The distinct diversity of the meq gene has been considered to be correlated with the enhanced virulence of MDV strains [19, 23]. The meq gene is thought to be associated with the evolution of MDV virulence, and non-synonymous mutations are frequently observed in meq genes among MDV strains [47]. To date, sequences of the meq genes from MDV strains from various countries have been reported [24–32]. In the present study, the UL–US regions of MDV strains were divided into two clusters, Eurasia and North America, as previously reported [47], whereas the meq genes of MDV strains from various regions formed three clusters, North America, other regions including Eurasian countries, and L-meq. Unfortunately, most of the whole genome sequences of MDV strains for which meq genes were classified into the L-meq cluster, in the phylogenetic analysis of the meq gene, were not available. However, the L-meq genes of strains 814 and CU-2, which were classified into Eurasia and North America groups, respectively, possess an insertion in the transactivation domain, and therefore, these L-meq genes seemed to be classified into the L-meq cluster in this phylogenetic analysis. Thus, the meq gene and UL–US region seem to indicate similar phylogeny, except for the appearance of the L-meq cluster. The meq gene of Kgs-c1, for which the sequence is frequent among Japanese MDV strains, was classified into the cluster of other regions. In addition, the UL–US region of Kgs-c1 was classified into the Eurasian cluster. Taken together, the genetic characteristics of MDV strains distributed in Japan might be largely classified into the Eurasian cluster, similar to that observed with Kgs-c1. However, some meq genes from Japanese isolates were classified into the North American cluster. Therefore, it is possible that MDV strains with genetic characteristics similar to those of US strains are also present in Japan.
A deletion of amino acid sequences was found in MDV013 (glycoprotein L)/MDV013.5 (MHC class II beta chain binding protein) and MDV056 (probable membrane protein). The deletion in MDV013/MDV013.5 was also found in some highly virulent strains [19]. However, this deletion does not seem to be correlated with increased virulence [58, 59]. The Chinese and European strains showed a deletion in the same region of MDV056 [33, 40–47]. However, this deletion was also found in the vaccine strain 814 and therefore, it might not affect virulence, although its correlation with MD pathogenesis is unclear.
The historical background related to the introduction of vaccines in each country is different, and this seems to be correlated with differences in the evolution of MDV genomes in each country [47]. In Japan, HVT was initially approved for protection from MD in 1972. A few years after the initial introduction of HVT, field outbreaks were sporadically observed in HVT-vaccinated chickens. Therefore, other types of vaccines, specifically CVI988 in 1985 and a bivalent vaccine comprising CVI988 and HVT in 1988, were approved. Thereafter, CVI988 and multivalent vaccines including CVI988 have been widely used to prevent MD occurrences in poultry farms in Japan. In Europe, HPRS-16, which is an MDV strain that was originally isolated in the UK, was initially used as a live-attenuated vaccine [2]. Later, a vaccine derived from an attenuated MDV strain, CVI988, was used; currently, this vaccine is being used globally [2]. In contrast, in the USA, HVT was initially developed as a live vaccine [2]. In addition, a naturally non-pathogenic strain, SB-1, was isolated in the USA in the late 1970s, and has been used as a bivalent vaccine to enhance the vaccine efficacy of HVT [2]. The historical background related to the use of vaccines between Europe and the USA is thus different, and the background of Japan is closer to that of Europe. Thus, the use of vaccines could induce the evolution of MDV strains in each country, and Japanese strains seemed to develop genetic characteristics similar to those of European strains.
The UL36 protein, a large tegument protein encoded by a member of Herpesviridae, is known to form the innermost layer of the complex protein scaffold between the capsid and envelope [60]. MDV encodes a ubiquitin-specific protease as part of the N-terminal region of the UL36 protein, similar to that observed in other known herpesviruses [61]. The UL36 protein was found to be correlated with the tumorigenic activity and replication of MDV via the deubiquitinase activity of the ubiquitin-specific protease [61, 62]. In contrast, MDV encodes unique repeat sequences at the C-terminal region [57]. Phylogenetic analysis revealed that the UL36 gene of Kgs-c1 was classified into the North American cluster, unlike those of the meq gene and UL–US region. In addition, Kgs-c1 exhibited unique sequences in the coding regions at the 5′ regions of the UL region. Thus, Kgs-c1 might have undergone evolutionary processes different from those of other pathogenic strains, although Kgs-c1 is closer to Chinese and European strains than to US strains in terms of the whole genome sequence.
For the phylogenetic analysis, methods based on the maximum likelihood principle have been often applied, as their accuracy is generally higher than that of other methods. However, these methods require much longer time, and some factors, such as the number of nucleotides, the number of sequences, and the model used for analysis, often affect accuracy [56]. Therefore, when using these methods, optimization should be considered. In addition, many empirical studies have indicated that other methods, including the minimum evolution principle, provide similar phylogenetic inference by applying the bootstrap test [56]. A previous study reported that in metagenomic analysis, phylogenetic classification using the novel method (PhyClass) based on the minimum-evolution principle was as efficient as that with the maximum likelihood methods [63]. Therefore, in the present study, the genetic characteristics of the Kgs-c1 genome were analyzed using the minimum-evolution method. However, its accuracy might be less than those using the best existing maximum likelihood methods with optimized models. Therefore, other approaches should be applied to analyze the phylogeny of Kgs-c1 more accurately, and assess the biases by the minimum-evolution method.