BK virus (BKV) infection and end-organ disease remains a formidable challenge to the hematopoietic cell transplant (HCT) and kidney transplant fields. As BKV-specific treatments are limited, immunologic-based therapies may be a promising and novel therapeutic option for transplant recipients with persistent BKV infection. Here, we describe a whole-genome, deep sequencing methodology and bioinformatics pipeline that identifies BKV variants across the genome and at BKV-specific HLA-A2, HLA-B0702, and HLA-B08 restricted CD8 T-cell epitopes. BKV whole genomes were amplified using long-range PCR with four inverse primer sets and fragmentation libraries were sequenced on the Ion Torrent PGM. An error model and variant calling algorithm were developed to accurately identify rare variants. 65 samples from 18 pediatric HCT and kidney recipients with quantifiable BKV DNAemia underwent whole-genome sequencing. Limited genetic variation was observed. The median number of amino acid variants identified per sample was 8 (range 2-37, interquartile range 10), with the majority of variants (77%) detected at a frequency of less than 5%. When normalized for length, there was no statistical difference in the median number of variants across all genes. Similarly, the predominant virus population within samples harbored T-cell epitopes similar to the reference BKV strain that was matched for BKV genotype. Despite the conservation of epitopes, low-level variants in T-cell epitopes were detected in 77.7% (14/18) of patients. Understanding epitope variation across the whole genome provides insight into the virus-immune interface and may help guide the development of protocols for novel immunologic-based therapies.
View details for DOI 10.1128/JCM.01385-15
View details for PubMedID 26202116