There are no methods for culturing PRV-1; the majority of genome sequences have been assembled from direct sequencing of infected tissues. This approach is limited to samples which have high PRV-1 loads. Like many fish viruses, PRV-1 loads in wild salmonids and environmental substrates (e.g. seawater) are typically too low for successful direct sequencing.
In 2017, Quick et al. developed a multiplex PCR method for targeted enrichment of Zika virus genomes and optimized library preparations for MinION and Illumina sequencing. They obtained full genome sequences from clinical samples with as few as 50 genome copies. We have adapted these techniques and applied them to obtain PRV-1 genomes directly from tissue and seawater samples.
PRV-1 genome sequences from British Columbia were used as reference to design 36 sets of overlapping primer pairs using Primal Scheme software (http://bioinformatics.uni-muenster.de/tools/nanopipe2/index.hbi?lang=en).
We routinely sequence 24-48 barcode-separated samples per flow cell and obtain 2-3 million reads for each 12-sample run. At this depth of sequencing we can generate accurate near full-length genome sequences from samples with as low as 10 PRV-1 copies per µl of sample template. Samples which had received improper handling or had been stored under suboptimal conditions for long periods of time (months to years) often had significant loss of amplification for some primer sets resulting in low read depth across portions of the genome. Nevertheless, even these mistreated samples typically yielded acceptable sequence coverage for the majority of the genome. The PRV-1 consensus sequences generated using the multiplex-MinION protocol are near-perfect matches (>99.9%) to lllumina consensus sequences from the same samples providing read coverage is greater than 20 reads per position.
We have obtained large numbers of PRV-1 genome sequences from fish and environmental samples, as well as recovered partial genome sequences from 43 year-old histology samples. Our plan is to integrate PRV-1 sequence data, epidemiological data and data from hydrodynamic modeling to determine transmission pathways within and between farmed and wild salmon populations over a range of temporal and spatial scales. Over the longer term our goal is to develop robust methods and analytical procedures that are suitable for application to a wide range of infectious agents in aquatic environments.