Genome Sequences of Ambystoma Tigrinum Virus Recovered during a Mass Die-off of Western Tiger Salamanders in Alberta, Canada

Complete genome sequences of six Ambystoma tigrinum viruses (ATV) were determined directly from tail clips of western tiger salamanders (Ambystoma mavortium) from 2013 (high-mortality year) and 2014 (low-mortality year) in Alberta, Canada. The genome lengths ranged from 106,258 to 106,915 bp and contained 108 open reading frames encoding predicted proteins larger than 50 amino acids.

A mbystoma tigrinum virus (ATV), a virus with a double-stranded DNA (dsDNA) genome of ϳ106 kb (1), and other emerging viruses of the genus Ranavirus and family Iridoviridae affect a wide range of amphibians, reptiles, and fish (2,3). To allow identification of candidate pathogenicity markers, DNA was extracted from the tail clips of tiger salamanders collected from Livingston Lake in Southern Alberta, Canada, during massive die-offs in 2013 with close to 100% mortality and during a year with almost no observed mortality, 2014. Archived DNA from samples positive for ATV by insulated isothermal PCR (iiPCR), which utilizes a temperature gradient instead of thermal cycling (4), was used to determine the full viral genome in five specimens collected in 2013 and one specimen collected in 2014 (Table 1). DNA was quantified using the Qubit 2.0 fluorometer (Thermo Fisher) and normalized to a 5 ng/l concentration with resuspension buffer (RSB) from the TruSeq Nano DNA kit for NeoPrep (Illumina) in a final volume of 15 l. DNA was sheared to ϳ550-bp fragments using an M220 ultrasonicator (Covaris) and processed on an Illumina NeoPrep instrument for sequence library preparation. Normalized sequencing library pools were sequenced on an Illumina MiSeq instrument using a V3 flow cell and a 600-cycle reagent cartridge (Table 1).
Following sequencing, the full viral genomes were generated by trimming raw reads using Trimmomatic (5) and then using SPAdes (v.3.9.1) (6) for de novo assembly. (from a total of 2,711,354 reads), which resulted in only a partial genome using SPAdes in metagenomics mode (metaSPAdes). Therefore, the sample was resequenced using probe capture enrichment (7). Resequencing was performed on an Illumina MiSeq instrument using a V2 flow cell and a 300-cycle reagent cartridge ( Table 1). The complete genome was assembled after filtering the reads using drVM (8) and then running the resulting Ranavirus reads through SPAdes in careful mode.
Raw sequencing reads from each sample were mapped back to the assembled genomes in Geneious (v.3.11.4) (9) and manually corrected for errors in the assemblies.
Prior to read mapping, the assembled genomes were circularized in Geneious to observe that the reads on the ends wrapped around the genome, therefore confirming their completeness. The complete genomes were searched for open reading frames (ORFs), using the Geneious ORFfinder, with a minimum length of 150 nucleotides. Coding sequences (CDS) from 15 complete ATV genomes imported from NCBI (23 June 2017) were transferred to the six samples using the Geneious Live Annotate and Predict tool with a 65% nucleotide identity threshold. An ORF and a CDS at the same position were considered a predicted protein. The minimum ORF length was then lowered to 48 nucleotides to find the remaining smaller CDS. GeneMarkS (10) and a local version of NCBI ORFfinder (11) were used to confirm the ORFs annotated in Geneious. Amino acid sequences found by the NCBI ORFfinder were run through the blastp command in DIAMOND (12) (Table 1), with the ORFs for the low-mortality-year samples being larger than those from the high-mortality-year samples. These differences may impact pathogenicity; however, further experimentation is required to confirm this.

ACKNOWLEDGMENTS
This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors. M.N. and S.G. were partially supported by funding from the Genomics Research and Development Initiative WIN-A-1408 and Ecobiomics projects, respectively.
We acknowledge Stephanie Reimer for assistance with field collection, the Brinkman family for access to Livingston Lake, and Amber Papineau and Mat Fisher for technical assistance.
We declare no conflict of interest. O.L. and C.G. conceived the study. S.G. prepared the next-generation sequencing (NGS) library. M.N. and O.L. analyzed the data. All authors contributed to the manuscript.