High-Throughput Sequencing of EBV Genomes Reveals Evidence for Strong Positive Selection Pressure From the Host Immune System.

Abstract

Around 90% of the adult population worldwide is Epstein-Barr Virus (EBV) seropositive and in most cases infection is controlled by host immunity. However, EBV is also linked to a number of malignancies including post-transplant lymphoproliferative disorder. Despite the prevalence of this virus, little is known about its genome-wide diversity, which could provide key insights into how the virus evolves in response to immune system pressure and the factors that determine whether EBV infection leads to disease. We characterized the genomic diversity of EBV by using a publically available dataset containing high throughput sequencing reads from EBV-infected cell lines. Following alignment of reads to the B95-8 reference genome, the GATK SNP calling module was used to determine sites that differed from the reference. This was followed by calculation of the average diversity (as quantified by Shannon entropy) over a sliding window across the genome. Overall, over 3,000 sites were found to be variable, 1400 of which were in protein-coding regions, and 796 were non-synonymous variants, representing a dS/dN ratio of 1.24. This suggests, as expected, that the EBV genome is under strong positive selection. The most diverse annotated genes as measured by SNP density were EBNA3c, LMP1, and EBNA3b, all genes involved in the latent phase of the viral lifecycle. This high level of diversity provides evidence consistent with tetramer-staining studies showing these genes are immunodominant and thus under strong selective pressure. At a larger scale, we identified several hotspots of diversity both within and outside of protein-coding regions that may be used in the future to track or elucidate the EBV family tree both within and between individuals. We are currently proceeding with sequencing transplant patient samples over time to analyze the effect the immune response has on viral diversity and connect variants with clinical outcomes.