Unbiased, next-generation sequencing for the characterization of Citrus tristeza virus populations

Abstract

A high-throughput sequencing pipeline to characterize Citrus tristeza virus isolates was developed. Three alternative viral templates (total RNA, double-stranded RNA and virus particles) were first tested on a single, previously characterized GFMS12 sub-isolate for their enrichment qualities, and combined with random RT-PCR amplification were subjected to Illumina paired-end sequencing. Double-stranded RNA was found to be most useful and was selected for further characterization of additional isolates (glasshouse-kept and field-derived). A novel South African genotype, named CT-ZA3 was assembled de novo and shown to be the dominant component in all GFMS12 sub-isolates tested. Genotype distributions within field-derived isolates collected from commercial orange (Citrus sinensis) orchards revealed a mixed infection status, dominated by a resistance breaking (RB)-like component (Tai-SP) coupled with a minor, VT-like (mild) (Kpg3) component. Based on read mapping patterns from field isolates, it is further suggested that two previously unknown recombinants may be present: a SP/Kpg3 and HA16-5/Kpg3 combination. This study underlined the effectiveness of next-generation sequencing for genotype discovery as well as whole-genome characterization of CTV isolates to a level of detail previously unreachable with classical methods such as SSCP and Sanger sequencing of multiple clones.