Complete Genome Sequences and Pathogenicity Analysis of Two Red Sea Bream Iridoviruses Isolated from Cultured Fish in Korea

Abstract

In Korea, red sea bream iridovirus (RSIV), especially subtype II, has been the main causative agent of red sea bream iridoviral disease since the 1990s. Herein, we report two Korean RSIV isolates with different subtypes based on the major capsid protein and adenosine triphosphatase genes: 17SbTy (RSIV mixed subtype I/II) from Japanese seabass (Lateolabrax japonicus) and 17RbGs (RSIV subtype II) from rock bream (Oplegnathus fasciatus). The complete genome sequences of 17SbTy and 17RbGs were 112,360 and 112,235 bp long, respectively (115 and 114 open reading frames [ORFs], respectively). Based on nucleotide sequence homology with sequences of representative RSIVs, 69 of 115 ORFs of 17SbTy were most closely related to subtype II (98.48–100% identity), and 46 were closely related to subtype I (98.77–100% identity). In comparison with RSIVs, 17SbTy and 17RbGs carried two insertion/deletion mutations (ORFs 014R and 102R on the basis of 17SbTy) in regions encoding functional proteins (a DNA-binding protein and a myristoylated membrane protein). Notably, survival rates differed significantly between 17SbTy-infected and 17RbGs-infected rock breams, indicating that the genomic characteristics and/or adaptations to their respective original hosts might influence pathogenicity. Thus, this study provides complete genome sequences and insights into the pathogenicity of two newly identified RSIV isolates classified as a mixed subtype I/II and subtype II.

1. Introduction

The virus species infectious spleen and kidney necrosis virus (ISKNV) (genus Megalocytivirus, family Iridoviridae) causes red sea bream iridoviral disease (RSIVD), which has a high mortality rate, in more than 30 susceptible freshwater and marine fish species [1]. According to the World Organization for Animal Health, it is a major fish disease [2]. Phylogenetic analyses based on major capsid protein (MCP) or adenosine triphosphatase (ATPase) genes have shown that the species can be classified into three major genotypes: red sea bream iridovirus (RSIV), ISKNV, and turbot reddish body iridovirus (TRBIV) [3]. The RSIV and ISKNV types can each be further categorized into two subtypes (I and II) [3]. Since the first outbreak of an RSIV-type infection among red sea breams (Pagrus major) in Japan in 1990 [4], RSIVs have been the predominant genotypes detected in marine fish in East Asian countries, including Korea [5,6,7]. In China, ISKNV and TRBIV types were first isolated from mandarin fish (Siniperca chuatsi) in 1998 [8,9] and from turbot (Scophthalmus maximus) in 2002, respectively [10]. In Korea, two genotypes of Megalocytivirus have been reported as endemic and have been taxonomically classified as RSIV [6,7,11] and TRBIV types [12]. Of note, RSIV subtype II has been identified as the major causative pathogen of endemic RSIVD in cultured marine fish in Korea [5].

Recently, an ISKNV/RSIV recombinant type was isolated from red sea bream (Pagrus major) in Taiwan, known as RSIV-Ku [13]. Its genome shares a high degree of homology with ISKNV-type viruses, except for specific nucleotide sequences that are closely related to RSIV-type viruses, implying that RSIV-Ku is a natural recombinant of ISKNV- and RSIV-type viruses [13]. Moreover, RSIV SB5-TY from a diseased Japanese seabass (Lateolabrax japonicus) in Korea is believed to be a genetic variant of RSIV-type viruses based on sequence difference in MCP and ankyrin repeat domains [5]. The emergence of genetic recombinants or variants of Megalocytivirus is a possibility, especially in RSIVD-endemic regions, such as Korea. Therefore, pathogenicity and complete genome sequence analyses of isolates in susceptible hosts are crucial for epidemiological studies, such as studies of source tracking and virus transmission.

In this study, we determined the complete genome sequences of two RSIVs identified in two cultured marine fish species (Japanese seabass and the rock bream (Oplegnathus fasciatus) in Korea, and analyzed insertion/deletion mutations (InDels). In addition, to evaluate their pathogenicity, a challenge test was performed on rock breams, which are known to be highly susceptible to RSIV infection.

2. Materials and Methods

2.1. Viral Culture

Primary cells derived from the fins of rock breams were grown in the L-15 medium supplemented with 10% fetal bovine serum (Performance Plus; Gibco, Grand Island, NY, USA) and 1% antibiotic-antimycotic solution (Gibco), as described by Lee et al. [14]. Briefly, caudal fin tissue was collected from juvenile rock bream (bodyweight, 5.4 ± 0.8 g), minced into small pieces (approximately 1 cm³), and then washed with phosphate-buffered saline (PBS). Cells treated with a 0.25% trypsin-EDTA solution (Gibco) at 20 °C for 1 h were filtered through a cell strainer (pore size: 70 μm; Falcon, NY, USA). Filtered cells were collected via centrifugation at 500× g for 10 min at 4 °C and were then resuspended in the culture medium and seeded in 25 cm² tissue culture flasks. The primary cells were incubated at 25 °C, and the medium was replaced daily. The cells were subcultured (split ratio: 1:2) when monolayer cells reached >90% confluence.

Tissue samples (spleen and kidney, 50 mg) were collected from diseased Japanese seabass in Tongyeong and rock bream in Goseong in 2017. To identify RSIV infection, real-time polymerase chain reaction (PCR) [15] was carried out. Briefly, each 20 μL real-time PCR mixture contained 1 μL of DNA, which was extracted using the yesG^TM Cell Tissue Mini Kit (GensGen, Busan, Korea), 200 nM each primer and probe (Table A1), 10 μL of the 2× HS Prime qPCR Premix (Genet Bio, Daejeon, Korea), 0.4 μL of the 50× ROX dye, and 5.6 μL of nuclease-free water. Amplification was performed using a StepOne Real-time PCR system (Applied Biosystems, Foster City, CA, USA) under the following conditions: 95 °C for 10 min, followed by 40 cycles of 94 °C for 10 s (denaturation) and 60 °C for 35 s (annealing and extension). Tissue samples that were RSIV-positive, as determined by real-time PCR, were used as the viral inoculum.

Viral infection (each tissue homogenate, 10 mg/mL) was induced in 75 cm² tissue culture flasks (Greiner Bio-one, Frickenhausen, Germany) containing monolayers of primary cells at passage 15. RSIV-infected cells were propagated at 25 °C for 7 days in L-15 medium containing 5% fetal bovine serum and 1% antibiotic-antimycotic solution. After the appearance of the cytopathic effect (rounded cells; Figure A1), the infected cells were collected and subjected to three freeze-thaw cycles. After centrifugation at 500× g for 10 min, the virus-containing supernatants were collected and stored at −80 °C until use. The cultured RSIVs were designated as 17SbTy and 17RbGs based on the sampling year, common name of the fish, and sampling site (i.e., 2017, Japanese seabass, Tongyeong and 2017, rock bream, Goseong).

2.2. Phylogenetic Analysis

For genotyping, genes encoding MCP and ATPase were amplified with the primers listed in Table A1 and sequenced using an ABI 3730XL DNA Analyzer (Applied Biosystems, CA, USA) by Bionics Co. (Seoul, Korea). Then, the MCP and ATPase gene sequences were quality-checked by base-calling using ChromasPro (ver. 1.7.5; Technelysium, Tewantin, Australia). Each sequence was identified using Nucleotide Basic Local Alignment Search Tool (BLASTn; https://blast.ncbi.nlm.nih.gov/Blast.cgi). Contigs were generated using the ChromasPro and aligned using the ClustalW algorithm in BioEdit (ver. 7.2.5). Phylogenetic trees were generated by the maximum likelihood method via the Kimura two-parameter (K2P) model with a gamma-distribution and invariant sites (K2P + G4 + I) using MEGA (ver. 11). The MCP and ATPase genes of epizootic haematopoietic necrosis virus (GenBank accession no. FJ433873) were used as outgroup in the phylogenetic analyses. Support for specific genotypes of the RSIVs were determined with 1000 bootstrap replicates (≥70%).

2.3. Determination of Complete Genome Sequences by Next-Generation Sequencing

Viral nucleic acids were extracted from gradient-purified virions using the QIAamp MinElute Virus Spin Kit (Qiagen, Hilden, Germany). Next, 1 μg of the extracted DNA was employed to construct sequencing libraries using the QIAseq FX Single Cell DNA Library Kit (Qiagen). Sequencing libraries of 17SbTy and 17RbGs were constructed, with average lengths of 648 bp and 559 bp, respectively. The quality of the libraries was evaluated using the Agilent High Sensitivity D 5000 ScreenTape System (Agilent Scientific, CA, USA), and the quantity was determined using a Light Cycler Real-time PCR system (Roche, Mannheim, Germany). The high-quality libraries (300–600 bp) were sequenced (pair-end sequencing, 2 × 150 bp) by G&C Bio Co. (Daejeon, Korea) on the Illumina HiSeq platform (Illumina, San Diego, CA, USA). To assess the quality of the sequence data, FastQC [16] and MultiQC [17] were employed. Low-quality sequences (base quality <20) and the Illumina universal adapters were trimmed from the reads using Trim-Galore software (ver. 0.6.1; https://www.bioinformatics.babraham.ac.uk/projects/trim_galore, accessed on 21 June 2020). High-quality reads were mapped and assembled into contigs using gsMapper (ver. 2.8). Nucleotide errors in the reads were corrected with the Illumina sequencing data using Proovread [18].

2.4. Complete Genome Sequence Analysis

2.4.1. Construction of a Circular Map

The composition, structure, and homologous regions of the genomic DNA were analyzed and circular map was generated using the cgview comparison tool [19]. Coding regions were classified according to a clusters of orthologous groups (COG) analysis. To determine COG categories, a comparative analysis was performed based on the proteins encoded in 43 complete genomes representing 30 major phylogenetic lineages described by Tatusov et al. (1997 and 2001) [20,21] using the COG program on the National Center for Biotechnology Information (NCBI) website (http://www.ncbi.nlm.nih.gov/COG, accessed on 12 April 2021). The genes were categorized in accordance with their functional annotations.

2.4.2. Gene Annotation and Open Reading Frame (ORF) Analysis

To identify putative ORFs, the full-length genome sequences of 17SbTy and 17RbGs were annotated using Prokka (ver. 2.1). ORFs were predicted using NCBI ORFfinder (https://www.ncbi.nlm.nih.gov/orffinder, accessed on 15 April 2021), and then the amino acid sequences of the putative ORFs were checked by Protein BLAST (BLASTp; https://blast.ncbi.nlm.nih.gov/Blast.cgi, accessed on 16 April 2021). Nucleotide sequence homologies of the putative ORFs of 17SbTy with those of 17RbGs and representative megalocytiviruses, i.e., Ehime-1 (GenBank accession no. AB104413; RSIV subtype I and the ancestral strain of RSIVD) [22], ISKNV (GenBank accession no. AF371960) [8], and TRBIV (GenBank accession no. GQ273492)] [23] were determined using BLAST (https://blast.ncbi.nlm.nih.gov/Blast.cgi, accessed on 10 May 2021). Furthermore, to analyze genetic relatedness among viruses in Iridoviridae, amino acid sequences of 26 conserved genes [24,25] were retrieved from NCBI GenBank. A phylogenetic tree based on the deduced amino acid sequences of 26 concatenated genes was constructed by the maximum likelihood method with the LG model and gamma-distributed rates with invariant sites (LG + G4 + I) [26] using MEGA (ver. 11.). Support for specific genera of iridoviruses was determined with 1000 bootstrap replicates (≥70%).

2.4.3. Analysis of InDels in RSIVs

To identify InDels in coding regions, the nucleotide sequences of 17SbTy and 17RbGs were compared with those of the ancestral RSIV (Ehime-1 isolated from a red sea bream in Japan in 1990; RSIV subtype I) [22] and an RSIV genome previously reported in Korea (RBIV-KOR-TY1 isolate found in a rock bream in 2000; RSIV subtype II; GenBank accession no. AY532606) [27]. Genomic sequences coding for functional proteins were aligned using the ClustalW algorithm in BioEdit (ver. 7.2.5), and InDels in the coding regions were detected.

2.5. Pathogenicity of the Two RSIV Isolates in the Rock Bream

Healthy rock bream (body length: 8.75 ± 1.95 [mean ± SD]; body weight: 6.79 ± 4.16 g) were obtained from an aquaculture farm in Geoje, Korea, after confirming that they were RSIV-free by PCR, as described in the Manual of Diagnostic Tests for Aquatic Animals for RSIVD [2,28], and by real-time PCR [15] (Table A1). The fish were acclimated in a 500 L aqua tank at 25.0 ± 0.5 °C for 2 weeks and were fed a commercial diet once daily. Each day, 50% of rearing water was replaced with temperature-adjusted (25 °C) fresh seawater. To prepare a viral inoculum, viral genome copy numbers of cultured 17SbTy and 17RbGs were determined by real-time PCR [15] with a standard curve constructed using the serial dilutions of a plasmid containing the MCP gene of 17RbGs. In a challenge test, each fish group was intraperitoneally injected with 0.1 mL of 17SbTy (n = 18; 10⁴ viral genome copies per fish), 17RbGs (n = 18; 10⁴ viral genome copies per fish), or PBS (n = 18; a negative control). After the viral challenge, the fish were maintained at 25.0 ± 0.5 °C in 30 L aqua tanks for 3 weeks, with 50% of water exchanged daily. DNA was extracted from the spleen tissue of dead fish, and RSIV infection was confirmed by real-time PCR. Survival rates were compared among the experimental groups by the log-rank test using GraphPad Prism (ver. 8.4.3.). Statistical significance was set at p-values < 0.05. Furthermore, the nucleotide sequences around four InDels in coding regions (ORFs 014R, 053R, 054R, and 102R on the basis of the 17SbTy isolate) were compared between cell-cultured isolates and viruses from RSIV-infected fish. DNA was extracted from three fish in each experimental group, and PCRs were carried out with each specific primer set (Table A1). Each 20 μL PCR mixture contained 1 μL of DNA (extracted using the yesG^TM Cell Tissue Mini Kit; GensGen, Korea), 500 nM each primer, 10 μL of the 2× ExPrime Taq Premix (Genet Bio, Daejeon, Korea), and 7 μL of nuclease-free water. Amplification was performed on an Alpha Cycler 1 machine (PCRmax, Staffordshire, UK) under the following conditions: 95 °C for 10 min, followed by 35 cycles at 94 °C for 30 s (denaturation), 55 °C for 30 s (annealing), and 72 °C for 60 s (extension). The amplicons were sequenced using the ABI 3730XL DNA Analyzer (Applied Biosystems) by Bionics Co. Contigs were assembled using ChromasPro (ver. 1.7.5) and aligned using the ClustalW algorithm in BioEdit (ver. 7.2.5).

3. Results & Discussion

The complete genome sequences of two RSIV isolates collected from representative fish susceptible to RSIVD (17SbTy from a Japanese seabass and 17RbGs from a rock bream) in Korea were investigated, and a comparative analysis of the pathogenicity of the isolates was performed. A phylogeny based on genes encoding MCP and ATPase revealed that 17RbGs belongs to RSIV subtype II, which has been the predominant genotype in marine fish in Korea since the 1990s [5]. Notably, 17SbTy grouped with subtype I or II of RSIV in phylogenetic analyses based on MCP or ATPase, respectively (Figure 1). Comparisons of 17SbTy with Ehime-1 (ancestral RSIV subtype I) and 17RbGs (RSIV subtype II), showed 99.63% and 98.24% identity for the MCP gene and 99.03% and 100% identity for the ATPase gene, respectively. Golden mandarin fish iridovirus, an RSIV subtype I reported in Korea in 2016 [29], shares 99.9% sequence homology with Ehime-1 in both the MCP and ATPase genes. Unlike golden mandarin fish iridovirus, 17SbTy was classified as a mixed RSIV subtype (subtype I/II).

Figure 1. Phylogenetic trees based on the complete nucleotide sequences of the (a) major capsid protein gene (MCP; 1362 bp) and (b) adenosine triphosphatase gene (ATPase; 721 bp) of two red sea bream iridovirus (RSIV) isolates (17SbTy and 17RbGs) collected from cultured fish in Korea. The phylogenetic trees were constructed using the maximum-likelihood method in MEGA (ver. 11). Bootstrap values were obtained from 1000 replicates, and the scale bar represents 0.05 nucleotide substitutions per site. The two RSIV isolates (17SbTy and 17RbGs) from this study are highlighted in bold and red color.

The complete genomes of 17SbTy (122,360 bp, GenBank accession no. OK042108), and 17RbGs (122,235 bp, GenBank accession no. OK042109) were similar in size to the genomes of most representative megalocytiviruses, RSIV (Ehime-1; 112,415 bp), ISKNV (112,080 bp), and TRBIV (110,104 bp), except for scale drop disease virus (GF_MU1; GenBank accession no. MT521409; 131,129 bp). The sequences were circularly permuted and assembled into a circular form, similar to most Megalocytivirus genomes (Figure 2). In addition, the G+C contents of the 17SbTy and 17RbGs genomes were 53.28% and 53.13%, respectively.

Figure 2. Circular genome maps of (a) 17SbTy (112,360 bp) and (b) 17RbGs (112,235 bp). From the inner ring to the outer ring, the first and eighth circles represented the genomic length (kbp) and nucleotide positions, respectively. The second and third circles show the G+C skew and G+C content, respectively. The fourth and fifth circles represent rRNA and tRNA genes on forward and reverse strands, respectively. The sixth and seventh circles indicate the functional categories of the protein-coding sequences in terms of clusters of orthologous groups (COG) on the forward and reverse strands, respectively.

In total, 115 and 114 putative ORFs were predicted in 17SbTy and 17RbGs, respectively (Table A2). The putative ORFs of 17SbTy (total length 104,868 bp, 93.3% of the genome) ranged in size from 111 to 3849 bp and encodes 36 to 1282 amino acid residues. Of the 115 ORFs, 70 were located on the sense (R) strand, and 45 were on the anti-sense (L) strand (Table A2). The putative ORFs of 17RbGs (total length 105,003 bp, 93.6% of genome) ranged in size from 111 to 4155 bp, encoding for 36 to 1384 amino acid residues. Of the 114 ORFs, 68 were located on the R strand and 46 were on the L strand. Of the annotated ORFs in 17SbTy (115 ORFs) and 17RbGs (114 ORFs), 43 (37.7%) and 42 (36.8%), respectively, could be assigned to a predicted structure and/or functional protein. The complete nucleotide sequences of 17SbTy and 17RbGs were closely related to rock bream iridovirus-C1 (RBIV-C1, GenBank accession no. KC244182) with identities of 99.56% and 99.69%, respectively. A comparison of the complete nucleotide sequences of 17SbTy and 17RbGs revealed 97.69% identity. In the ORFs of 17SbTy, nucleotide sequence identities were 87.99–100% with Ehime-1 (RSIV subtype I), 88.22–100% with 17RbGs (RSIV subtype II), 86.07–97.58% with ISKNV, and 80.25–99.66% with TRBIV (Table A2). Notably, the best matches for the nucleotide sequences of the 115 ORFs of 17SbTy were RSIV subtype II viruses (97.48–100% identity for 69 ORFs) and RSIV subtype I viruses (98.77–100% identity for 46 ORFs).

A total of 20 protein-coding genes in both 17SbTy (17.39%; 20/115 ORFs) and 17RbGs (17.54%; 20/114 ORFs) were annotated in the COG database, and these genes were assigned to nine functional groups (Table A3): (i) amino acid transport and metabolism; (ii) nucleotide transport and metabolism; (iii) translation, ribosomal structure, and biogenesis; (iv) transcription; (v) replication, recombination, and repair; (vi) signal transduction mechanisms; (vii) mobilome, prophages, transposons; (viii) general function prediction only; and (ix) function unknown. The nine functional groups identified in both 17SbTy and 17RbGs belonged to four major categories: metabolism, information storage and processing, cellular processes, and poorly characterized. Furthermore, both 17SbTy and 17RbGs harbored the 26 conserved genes that were shared by all members of the family Iridoviridae, including genes encoding enzymes and structural proteins involved in viral replication, transcriptional regulation, protein modification, and host-pathogen interactions [24,25]. The ORFs corresponding to these 26 core genes are listed in Table A4. A phylogenetic tree based on the concatenated amino acid sequences of the 26 conserved genes revealed that 17SbTy and 17RbGs can be assigned to the genus Megalocytivirus. Furthermore, 17SbTy was closely related to Ehime-1 (Figure 3).

Figure 3. Phylogenetic trees based on the deduced amino acid sequences of the 26 concatenated genes conserved for members of the family Iridoviridae. The tree was constructed by the maximum-likelihood method under the LG model and gamma-distributed rates with invariant sites (LG + G4 + I) in MEGA (ver. 11). The two RSIV isolates (17SbTy and 17RbGs) from this study are highlighted in bold and red color.

As described by Eaton et al. (2007) [24], several annotated genes within the family Iridoviridae contain frameshift mutations. InDels are a type of frameshift mutation that can affect the translation of a functional protein. The complete genome of 17SbTy showed 133 InDels when compared to the Ehime-1 and 17RbGs genomes (data not shown). Notably, although the genomes of several RSIVs, including 17SbTy, Ehimel-1, and RBIV, encode two functional proteins—an mRNA-capping enzyme (ORF 012R, positions 10,693–12,165 in the 17SbTy genome) and a putative NTPase I (ORF 013R, positions 12,205–14,853 in the 17SbTy genome)—17RbGs possesses only a single functional protein (ORF 012R, positions 10,690–14,844 in the 17RbGs genome; Figure 4). A frameshift mutation caused by a short InDel [a 6 bp deletion, including a stop codon (TGA) and an intergenic codon (CCT)] explained the difference in the total number of ORFs between 17RbGs (n = 114) and 17SbTy (n = 115; Figure 4 and Table A2).

Figure 4. Schematic representation of a deletion of the termination codon in ORF 012R of 17RbGs causing a frameshift mutation. The aligned sequences are genomes of 17SbTy, 17RbGs, and two representative RSIVs (Ehime-1 [RSIV subtype I] and RBIV-KOR-TY1 [RSIV subtype II]). The nucleotide sequences surrounded by blue dashed lines are coding regions. The termination and start codons are shown in red, and the deleted sequences in the intergenic region are highlighted in blue.

Among the InDel regions in 17SbTy identified in comparisons with the Ehime-1 and 17RbGs genomes, 18 regions contained >10 bp mutations, and only four InDels were identified in coding regions (ORFs 014R, 053R, 054R, and 102R in 17SbTy). Although two ORFs encode known functional proteins (ORF 014R, which is involved in DNA binding, and ORF 102R, which is a myristoylated membrane protein; Figure 5a,d), two additional ORFs (ORF 053R and 054R) have not yet been functionally characterized (Figure 5b,c). Of the InDels found in the ORFs known to encode functional proteins, a 27 bp deletion in a DNA-binding protein with an FtsK-like domain was identified in 17SbTy (ORF 014R), in 17RbGs (ORF 013R), and RBIV-KOR-TY 1 (ORF 058L), but not in Ehime-1 (ORF 077R; Figure 5a). The FtsK-like domain in spotted knifejaw iridovirus (an RSIV-type) [30] participates in host immune evasion by inhibiting transcriptional activities of NF-κB and INF-γ, indicating that the deleted sequences in the gene encoding a DNA-binding protein might affect viral replication and/or pathogenicity. Furthermore, ORF 102R of 17SbTy, located in the same region as ORF 575R in Ehime-1, encodes a myristoylated membrane protein, known as a viral envelope membrane protein of iridovirus, and its function may be conserved throughout the family Iridoviridae [31]. Thus, an InDel in the coding region of a viral membrane protein (a 30 bp deletion in ORF 101R of 17RbGs) may alter the regulation of viral entry into host cells at the onset of the infection cycle.

Figure 5. Schematic representation of insertion/deletion mutations (InDels) (>10 bp) in the coding regions as (a) ORF 014R, (b) ORF 053R, (c) ORF 054R and (d) ORF 102R based on the 17SbTy when compared with the genomes of 17RbGs and two representative RSIVs (Ehime-1 [RSIV subtype I] and RBIV-KOR-TY1 [RSIV subtype II]). Numbers indicate the positions of the InDels in the genome; white bars represent genome fragments, black bars denote insertions, and gray bars represent deletions.

No rock bream infected with 17RbGs survived 15 days post-injection, whereas 27.8% (5/18) of the 17SbTy-infected rock bream survived 21 days post-injection (Figure 6). The difference in survival rates between the 17SbTy- and 17RbGs-infected rock breams was significant (log-rank test, p < 0.001). The nucleotide sequences of the four InDel regions (ORFs 014R, 053R, 054R and 102R on the basis of the 17SbTy isolate) were identical in the cell-cultured isolates and viruses from dead fish (Figure A2). These results suggest that several of the genetic factors identified in the genomic analysis, including the InDels in coding regions, may influence virulence. Another noteworthy observation is that the apparent difference in virulence between the RSIV isolates may be due to adaptations to their respective original hosts (Japanese seabass for 17SbTy and rock bream for 17RbGs). Further molecular epidemiological studies, including analyses of RSIV replication and pathogenic determinants, are needed to elucidate the transmission of RSIV.

Figure 6. Survival rates (%) of rock breams after intraperitoneal injection with the two RSIV isolates (either 17SbTy or 17RbGs, 10⁴ genome copies per fish). Statistical analysis was performed by the log-rank test (* p < 0.05).

4. Conclusions

Phylogenetic trees based on genes encoding MCP and ATPase revealed that two RSIV isolates (17SbTy from a Japanese seabass and 17RbGs from a rock bream) can be classified as RSIV mixed subtype I/II and subtype II, respectively. According to complete genome analysis, these isolates (17SbTy, 112,360 bp; 17RbGs, 112,360 bp) have the genomic organization, G+C content, coding capacity, and conserved core genes typical of the species ISKNV. Notably, the best matches for the nucleotide sequences in the 115 ORFs of 17SbTy were RSIV subtype II (69 matching ORFs; 97.48–100% identity) and RSIV subtype I (46 matching ORFs; 98.77–100% identity). In comparison with RSIVs, 17SbTy and 17RbGs had InDels in ORFs 014R and 102R (based on the 17SbTy genome), encoding a DNA-binding protein and myristoylated membrane protein, respectively. The survival rates of rock breams infected with these isolates differed significantly, suggesting that the genomic differences between these viruses and/or adaptations to their respective original hosts may have altered their pathogenicity. Thus, the complete genome sequences of these RSIV isolates provide basic information for molecular epidemiology and are expected to provide insight into viral replication in general and the pathogenicity of these viruses in susceptible hosts in particular.