Isolation and characterization of the fall Chinook aquareovirus

RNA from the tissue culture supernatants were extracted with Trizol according to the RNA extraction miniprep protocol (DirectZol, Zymo Research) and screened using sets of degenerate aquareovirus pan-species RT-PCR primers (Table 1). For each test, known positive control RNA from isolates of Aquareoviruses A and B were included along with deioinzed water as a negative control. Forward and reverse primers as shown in Table 1 for either Aquareovirus A (1811F/1812R) or Aquareovirus B (1807F/1809R) were tested in separate reactions. A nested PCR without the reverse transcription step, reverse transcriptase or RNasin enzymes was conducted using internal primers (Table 1) for Aquareovirus A (1813F/1814R) or for Aquareovirus B (1808F/1810R). Because the CPE produced by the fall Chinook aquareovirus on the CHSE-214 cells appeared somewhat different from that in a companion flask of cells infected with an Aquareovirus B isolate from a different location, we performed another RNA extraction and RT-PCRs with primers 1807F/1809R and 1811F/1812R. Additionally, after 13 days post infection, a further RNA extraction and PCR run was tested in combination with several isolates of putative aquareoviruses from locations near the Glenwood Springs Hatchery. The RT-PCR was performed in 50uL reactions using 28uL deionized water, 10uL 5X PCR buffer, 3uL 25 mM MgCl2, 1uL 10 mM dNTP, 1uL of each 20uM each primer, 0.25uL GoTaq, and 0.5uL AMV reverse transcriptase, and 0.25uL RNasin with 5uL of RNA template. RT-PCR conditions were 50C 30 min; 95C 2 min; 30 cycles of 95C 30s, 50C 30s, 72C 1 min; followed by 72C for 7 min and a 15C hold. PCR products were visualized on an ethidium bromide stained 1.5% agarose gel.

For genome recovery, RNA from supernatant was extracted using the Zymo Viral RNA Kit (Zymo Research) and treated with Turbo DNAse I (Thermo Fisher) [9]. cDNA synthesis was performed using random hexamers with SuperScript III reverse transcriptase and second-strand synthesis was performed using Sequenase v2.0 (Thermo Fisher). Double-stranded DNA was cleaned using a Zymo DNA-5 Clean and Concentrator column (Zymo Research) and library preparation was performed using 0.4X volumes of Nextera XT (Illumina), 16 cycles of dual-indexed library amplification, purification with 0.75X Ampure XP beads (Beckman Coulter). The library was quantified on Qubit 3.0 and Agilent Bioanalyzer 2100 and sequenced using a portion of a 1x180bp sequencing run on an Illumina MiSeq [10].

Sequencing reads were adapter and quality trimmed using cutadapt, repaired using pairfq, and de novo assembled using SPAdes v3.8 [11]. The resultant contigs were aligned by megablast to the NT database and the unaligned contigs were aligned by DIAMOND to the NR database [12]. The alignment files were taxonomically assigned and visualized using MEGAN 2.0, while contigs were visualized in Geneious v9.1 and phylogenies were performed using MAFFT alignments and MrBayes for phylogenetic trees [13‐15].

Follow-up RT-PCR was performed to close a gap in the NS1 gene using NS1–1386R and NS1-374F primers (Table 2) and quarter-volumes of the Qiagen One-Step RT-PCR kit. 3’RACE was performed as described previously [16]. Briefly, post-DNAsed RNA was polyadenylated by using E. coli Poly(A) Polymerase (NEB). Poly(A) RNA was reverse transcribed using Super Script III RT with Q_T primer and 30 cycles of Phusion PCR with a T_m of 55 °C with the Q _outer primer and GSP1 (Table 2) [16]. Nested PCR was performed using 1uL product from the first round PCR with 35 cycles using the Q inner primer and associated GSP2. PCR products were visualized on a 1.5% agarose gel and bands were gel-extracted using the QIAquick Gel Extract kit and sequenced via Sanger sequencing.

After the appearance of initial CPE in the primary culture, three, 25 cm² flasks were seeded with CHSE 214 cells and allowed to incubate at 15 °C overnight with MEM containing 10% fetal bovine serum (MEM-10) until cells were 90–100% confluent. When the cell monolayer was confluent, growth medium was removed and 200 μL of primary culture was filtered through a 0.2 μM membrane into two of the three flasks. The negative control was 200 μL of MEM-10 into a third flask. Flasks were gently swirled and allowed to incubate for 1 h at 15 °C. After 1 h, 7 mL of MEM-5 was added to each flask. A 200uL aliquot for PCR was taken to quantify the starting inoculum. On days 1–11, flasks were observed for CPE and 200 uL aliquots were collected from each flask and frozen at −20 °C.

A quantitative PCR was designed based on alignment of the nucleotide sequences of selected segments of other aquareoviruses. Primers to conserved regions of the viral polymerase (VP2) were designed (Table 2) and a positive control amplicon for use in developing a standard curve was created using the Qiagen One-Step RT-PCR kit.

Double-stranded viral RNAs were extracted from culture supernatant using the Pure Link RNA mini Kit (Thermo Fisher) and eluted with 60uL RNAse free water then frozen at −20 °C until used. Reverse transcription was completed using SuperScript II (Thermo Fisher) following manufacturer’s instructions. cDNA was frozen at −20 °C until qPCR could be completed. Primers provided were optimized using the SYBR Green recommendation for a final primer concentration of 0.3uM. The control standard was quantified using a Qubit (Thermo Fisher) and a known stock solution prepared (10⁶ copies/uL) for use in creating a standard curve.

PCR master mix consisted of a 1X final concentration SYBR Green Master mix (Applied Biosystems), 0.3uM of the forward and reverse primer and 1uL of cDNA template for a final volume of 10uL/ rxn. Samples were run in triplicate using a 384 well plate in a Quant studio 6 Flex PCR machine (Applied Biosystems). The standard curve was prepared as a dilution series of 10⁶ to 10¹ copies/rxn.

During the initial incubation of samples at the WDFW laboratory, CPE appeared on day 28 in only 1 of 2 wells of CHSE-214 cells inoculated with 1 of the 13 pools of kidney/spleen samples (Fig. 1a), indicating the virus was present in the population at low prevalence and intensity of infection. Beginning on day 4, flasks of CHSE-214 cells inoculated with filtered cell culture fluid from the affected well developed CPE typical of most aquareoviruses consisting of plaque-like areas of syncytia (Fig. 1b). Viral stocks were prepared by filtering culture fluid into cryovials that were frozen at −80 °C and sent to the United States Geological Survey Western Fisheries Research Center for testing. RT-PCR assays using several pairs of aquareovirus pan-species primers failed to amplify the extracted RNA (Fig. 2). Subsequently, the unidentified isolate was submitted to the Department of Laboratory Medicine at the University of Washington for next-generation sequencing.

A total of 187,569 adapter-trimmed reads were recovered from the RNA-sequencing libraries from the original cell culture supernatant. De novo assembly of the RNA-sequencing reads yielded 42 contigs longer than 500 bp. BLASTN and DIAMOND alignment indicated the presence of a novel aquareovirus, as alignments to various aquareoviruses over all eleven segments were found among the top 100 contigs.

Because of the dsRNA nature of the reovirus genome, 3’RACE was performed to finish both 5′ and 3′ ends. The fall Chinook aquareovirus genome contained 11 segments totaling 23.3 kb, ranging in size from 3947 bp to 728 bp (Fig. 3a). 5′NTR length ranged from 11 to 30 bp while the 3′NTR length ranged from 30 to 157 bp. All segments began and ended with the canonical aquareovirus sequence (5′-GUUUA and UCAUC-3′, Fig. 3b). All segments contained a single ORF, except S7 which contained two overlapping ORFs, suggesting that the fall Chinook aquareovirus encodes for 12 proteins. In total, 39,633 of the 187,569 adapter-trimmed reads (21.1%) mapped to the finished fall Chinook aquareovirus genome.

The fall Chinook aquareovirus VP7 shares a 63% nucleotide identity with the Green River Chinook virus (KC588384) and 47% amino acid identity with the Coho salmon aquareovirus (AAC59609). Phylogenetic analysis of both the VP4 and VP7 segments revealed clustering of the fall Chinook aquareovirus within the species Aquareovirus B (Fig. 4). Of note, the fall Chinook aquareovirus demonstrates the most divergence of any new aquareovirus that does not constitute a new species based on the ICTV definition as well as the first complete genome sequence for an aquareovirus B species. Examination of the VP2 coding sequence revealed the canonical amino acid motifs for polymerase (DXXXXD, SG, and GDD) present in members of the Aquareovirus and Orthoreovirus genera.

Pairwise nucleotide identity between the Green River chinook aquareovirus and the fall Chinook aquareovirus ranged from 63% (VP7) to 73% (VP3). The existing pan-Aquareovirus B primers had 3 mismatches in each of the outer primers and 9 mismatches in the hemi-nested second round primer in the VP7 segment, consistent with their inability to amplify the fall Chinook aquareovirus, and leading to the design of a novel primer set for confirmation of isolates of species Aquareovirus B (Table 1).

Recovery of the full genome sequence of the fall Chinook aquareovirus allowed development of a qRT-PCR assay that was used to study viral growth kinetics in flasks containing monolayer cultures of CHSE-214 cells. Samples of culture fluid removed from the flasks at daily intervals showed a strong increase in virus titer during the initial incubation period (Fig. 5).