Alphaherpesviruses are widespread in the human population, with herpes simplex virus 1 (HSV1) and 2 causing oral and genital lesions, respectively, while varicella zoster virus (VZV) causes chicken pox and shingles 
. In the agricultural industry, a related veterinary alphaherpesvirus, pseudorabies virus (PRV), causes similar disease in swine and significant economic cost due to weight loss in infected adults and reproductive losses during pregnancy and suckling 
. As occurs with HSV and VZV, PRV infection has higher morbidity and mortality rates for neonates, with decreasing severity of disease as the age at onset of infection increases 
. PRV and VZV primarily infect via the respiratory mucosa, while HSV-1 primarily infects at the oral mucosa. VZV infection includes a viremic phase that yields widespread vesicular lesions, while PRV and HSV are usually non-viremic and spread predominantly by mucosal infection and neuronal innervation. These alphaherpesviruses are widespread in the population because of their tendency to infect neurons: they establish lifelong latency in the host peripheral nervous system. These latent neuronal infections may occasionally reactivate and spread back the mucosal surfaces where the infection initiated. After further replication, the viruses can spread to new hosts.
Among alphaherpesviruses, vaccines are available for VZV and PRV, but not HSV 
. Despite considerable effort and recent progress, no broadly effective vaccine candidates have yet emerged for HSV infection 
. The co-morbidities of HSV-1 and HSV-2 with human immunodeficiency virus (HIV), which include increased acquisition of HIV due to the inflammation and lesions caused by HSV infection, have added impetus to the search for a vaccine 
. PRV serves as a useful model for HSV pathogenesis and vaccine development, because of their similar infectious cycle and ability to infect a variety of animal models 
. In contrast, VZV has a more restricted tropism for human cells that complicates its study in animal models 
. The agricultural importance of PRV and relative ease of vaccine testing has led to the development of several PRV vaccine strains, whose genetic characteristics have been determined by mapping isolated genomic fragments and sequencing of select regions 
. Of note, the vaccine strain Bartha has a well-characterized deletion of several viral proteins that attenuates its virulence and also limits its spread in neurons, which led to its subsequent development as a tool for trans-neuronal tracing 
. Like several other early vaccine strains, PRV Bartha was attenuated by extensive passage in the laboratory, thus making the full discovery of its genome-wide mutations a priority 
. Because the only available PRV genome sequence to date is a mosaic of six strains 
, it has been difficult to discern whether mutations detected in PRV Bartha and other vaccine strains are unique or represent ordinary sequence diversity, i.e.
are found in other wild-type genomes 
. We therefore applied our recent success in using Illumina high-throughput sequencing (HTS) to obtain HSV-1 strain genomes to determining the sequence diversity in the PRV vaccine strain Bartha.
In addition to sequence polymorphisms, insertions, and deletions, another major class of variation between nucleic acid sequences lies in copy number variation, either of coding sequences or of repeated structural elements. Herpesvirus genomes have long been known to contain several sites with tandem short sequence repeats (SSRs) or reiterations 
. Variation in these elements has been described both within and between herpesvirus strains, but their functions were largely unexplored 
. SSRs can be transcription factor binding sites, chromatin insulators, protein folding motifs, or other regulatory elements 
. Recent studies have shown that SSR expansion and contraction, most likely through recombination or polymerase slippage, can generate phenotypic variation 
. A range of human diseases result from SSR expansion or contraction, including the transcriptional silencing of the gene FMR1 via an upstream SSR, which causes Fragile X syndrome, and the poly-glutamine tract expansion in huntingtin protein, which causes Huntington's disease 
. Limited explorations of repetitive elements in viral genomes suggest that SSRs in viral genomes likewise play functional roles 
. To explore SSR prevalence and function in herpesviruses, we initiated a global SSR assessment and comparison across viral species, as was recently done for a variety of fungal and bacterial pathogens 
. These data highlight the contribution of SSRs to overall sequence diversity in viruses, and through the presence of these elements in both coding and non-coding regions, suggest that viral SSRs may likewise have the potential to affect gene expression and protein functions.
We sequenced three widely-studied PRV isolates by HTS: the attenuated vaccine strain Bartha and the virulent strains Kaplan and Becker. This analysis reveals genome-wide sequence diversity between strains, both in the PRV proteome and also in many SSRs. Our comparison of protein coding sequences revealed that 46 of 67 PRV proteins have changes in the vaccine strain Bartha which are not found in the virulent Kaplan or Becker strains. We mapped homologous SSRs in all three strains and provide a comprehensive overview of inter-strain variation in SSR length. We compared the proportion of SSRs in PRV to those found in HSV-1, VZV, the human betaherpesvirus cytomegalovirus (HCMV) and gammaherpesviruses Epstein-Barr virus (EBV) and Kaposi's sarcoma-associated herpesvirus (KSHV), and Mimivirus. We find that SSRs are likely to be a common property of these large DNA viruses. Finally, we examined the limited number of polymorphic bases detected in these plaque-purified virus stocks, and tested the rate of polymorphism occurrence in purified and non-purified virus populations. These data on sequence variation in PRV strains expand our understanding of viral genome diversity and how attenuated strains lead to successful anti-viral vaccines.