Whole-Genome Analysis of a Novel Fish Reovirus (MsReV) Discloses Aquareovirus Genomic Structure Relationship with Host in Saline Environments

Aquareoviruses are serious pathogens of aquatic animals. Here, genome characterization and functional gene analysis of a novel aquareovirus, largemouth bass Micropterus salmoides reovirus (MsReV), was described. It comprises 11 dsRNA segments (S1–S11) covering 24,024 bp, and encodes 12 putative proteins including the inclusion forming-related protein NS87 and the fusion-associated small transmembrane (FAST) protein NS22. The function of NS22 was confirmed by expression in fish cells. Subsequently, MsReV was compared with two representative aquareoviruses, saltwater fish turbot Scophthalmus maximus reovirus (SMReV) and freshwater fish grass carp reovirus strain 109 (GCReV-109). MsReV NS87 and NS22 genes have the same structure and function with those of SMReV, whereas GCReV-109 is either missing the coiled-coil region in NS79 or the gene-encoding NS22. Significant similarities are also revealed among equivalent genome segments between MsReV and SMReV, but a difference is found between MsReV and GCReV-109. Furthermore, phylogenetic analysis showed that 13 aquareoviruses could be divided into freshwater and saline environments subgroups, and MsReV was closely related to SMReV in saline environments. Consequently, these viruses from hosts in saline environments have more genomic structural similarities than the viruses from hosts in freshwater. This is the first study of the relationships between aquareovirus genomic structure and their host environments.


Introduction
Reoviruses are non-enveloped, icosahedral, double-stranded RNA (dsRNA) viruses which have double-layered capsids and genomes consisting of 9 to 12 genome segments [1]. Genus Aquareovirus belonging to the family Reoviridae can infect a wide variety of aquatic animals, including crustacean, shellfish, and fish [2]. These viruses are capable of causing severe hemorrhagic disease in fish and syncytia in cell culture, and have been isolated from hosts in both freshwater and saline environments worldwide [3][4][5][6][7][8][9][10][11][12]. Viruses in genus Aquareovirus have 11 dsRNA segments (S1-S11) [2]. Recently, a fish reovirus (piscine reovirus, PRV), which is proposed as a tentative new member of the family Reoviridae, has been identified to have 10 dsRNA segments [13]. Most segments of aquareovirus have only one open reading frame (ORF), and the genome usually encodes seven major structural proteins (VP1-VP7) and five major nonstructural proteins.
Phylogenetic analysis has been an increasingly efficient tool to examine aquatic animal virus epidemiology, and to infer common ancestors and their host relationships [14][15][16][17][18]. Comparative sequence analysis shows interesting similarities and dissimilarities among some equivalent genome segments of the reoviruses [19,20]. A similar genomic structure occurs in a closely related species [21,22]. Initial attempts to examine genetic variation among aquareovirus isolates primarily used partial or full-length genome sequences [23]. The primary goal was to distinguish between different aquareovirus strains. In recent years, aquareoviruses were recognized with high genomic variability [24,25]. The expanding knowledge of aquareovirus genetics has led to a number of groups proposing sequence-based typing schemes. For example, the phylogenetic tree of RNA dependent RNA polymerase (RdRp) revealed evidence for genetic relatedness and genomic diversity of aquareoviruses [26,27]. The phylogenetic comparison of reovirus genetic segments allowed the identification of reoviruses into Spinareovirinae and Sedoreovirinae subfamilies [1,28].
Until now, more than 15 aquareovirus genomes have been completely sequenced [12], including turbot Scophthalmus maximus reovirus (SMReV), and grass carp reovirus strain 109 (GCReV-109), as previously reported by our laboratory [24,26], and largemouth bass Micropterus salmoides reovirus (MsReV) described in this article. They have been found by way of the molecular variation that occurs throughout reoviral genomes, which are correlated with geographical distribution, classification (e.g., Aquareovirus and Orthoreovirus), morphological features (e.g., turreted reoviruses Spinareovirinae and non-turreted reoviruses Sedoreovirinae) [16,29,30]. Genomic variability may also assist in determining viruses with distinct cytopathic and pathogenic properties (e.g., causing syncytia and inclusion body formation) [26,31]. The properties of viral genome such as its size and chemical composition are identified as major determinants of evolutionary rates [32]. Some putative genes and the probable functions of encoded proteins in aquareoviruses have been reported [6,25,27,[33][34][35]. Aquareovirus GCRV S4 encoding NS80 ensures its self-aggregation to form viral factories like structures (VFLS) and recruitment of viral proteins [31]. Aquareoviruses have been used as a model to understand the structural basis and pathogenesis of reovirus [26,31,36]. An unexpected challenge that has arisen from aquareoviruses is the viral genetic diversity, and the prevention and control of aquareoviruses remains largely unaddressed as they are known to be very limited in the viral pathogen-host fish. Although genome diversity has been reported among aquareoviruses [12,[24][25][26], there is finite information on the natural processes that contribute to genome diversity, and it is still unclear whether the genome structure of aquareovirus is reflective of the interrelationships between the virus and host environment.
To assess the genomic variability of aquareoviruses and the relationships between the viruses and their host environments, here we investigated the aquareovirus genomic structure relationship with hosts in saline environments based on new as well as previously published sequence information, with a comparison of equivalent genomic segments and phylogenetic analysis. Moreover, the phenotype associated with the pathogenicity of specific gene MsReV NS22 was tested by construction and expression of plasmid with deletions or mutations.

Virus Isolation, Electron Microscopy and Electrophoretic Analysis
Diseased largemouth bass were collected in Hubei province of China in May 2010. Liver, spleen and kidney tissues were sampled and homogenized as previously described [26]. The suspension was centrifuged at 2000 g for 30 min and then filtered through a sterile 0.45 µm filter (Millipore, Billerica, MA, USA). The filtered supernatant was inoculated into confluent monolayers of bluegill fry (BF-2), chinook salmon embryo (CHSE-214), epithelioma papulosum cyprini (EPC), fathead minnow (FHM), grass carp fins (GCF) and grass carp ovary (GCO) cell lines in TC199 medium containing 5% fetal bovine serum at 15˝C, 20˝C or 25˝C. Inoculated cell cultures were checked daily for cytopathic effects. The original viral isolate was adapted to cell culture through at least three passages on these cell lines. The optimal temperature for virus propagation was assayed by infection of GCF cell monolayers at 15˝C, 20˝C or 25˝C. Viral titers were measured on the basis of 50% tissue culture infective dose (TCID 50 ) mL´1 as described previously [10].
Virus particles were purified from cell culture-amplified virus stocks as described previously [26]. Purified virus particles were negatively stained with 2% (w/v) phosphotungstic acid, and then examined with the Hitachi HT-7700 electron microscope (Hitachi, Tokyo, Japan).
Virus dsRNA was extracted from purified virus particles using Trizol Reagent according to the manufacture's instruction (Invitrogen, Carlsbad, CA, USA). The extracted dsRNA was analyzed on a 15% polyacrylamide gel in 0.5ˆTris-borate/EDTA (TBE) buffer, and then visualized by sliver staining. Genomic dsRNA from Scophthalmus maximus reovirus (SMReV) and grass carp reovirus strain 109 (GCReV-109) maintained in our laboratory [24,26] were prepared and used as molecular mass size markers.

Viral Genome Sequencing
The cDNA from virus dsRNA was synthesized using the single-primer amplification technique [26]. Briefly, an oligodeoxyribonucleotide primer (TC1: 5 1 PO 4 -CCCGCCATCCTCACTTAGACT-NH 2 3 1 ) was ligated to both of the 3 1 ends of the dsRNA segments by T4 RNA ligase (TaKaRa, Dalian, China). After the reaction, dsRNA was denatured at 94˝C for 5 min in the presence of 15% dimethyl sulfoxide (DMSO) and then cooled rapidly on ice. RNA was then removed by adding NaOH and the cDNA was annealed at 65˝C overnight. The first strand cDNA of the genome segments were synthesized using M-MLV (Promega, Madison, WI, USA), and then purified by a Sephacryl S-400 spin column (Promega). The amplification of the cDNA was performed using the complementary primer (TC2: 5 1 AGTCTAAGTGAGGATGGCGGG 3 1 ). PCR products were electrophoresed on 1% agarose gels, and all visible bands were purified and ligated into the pMD18-T vector (TaKaRa). The positive clones were sequenced on an ABI 3730XL DNA analyzer (Sangon, Shanghai, China).

Sequence Analysis and Comparison
The nucleotide sequences and deduced amino acid sequences were analyzed using the EditSeq program (DNASTAR 5.0). Homology searches of nucleic acid and protein databases were performed using BLAST at the National Centre for Biotechnology Information server. Multiple sequence alignments were performed using Clustal X 1.83 program, and sequence identities were calculated using the Clusta W method in the MegAlign program (DNASTAR 5.0). Transmembrane helices were predicted using TMHMM 2.0 [37]. The coiled regions in MsReV NS87 protein were predicted using the COILS Server (http://embnet.vital-it.ch/software/COILS_form.html). The equivalent genome segments and proteins between MsReV and two other representative aquareoviruses, SMReV and GCReV-109, were analyzed and shown in a schematic diagram.
GCF cells grown in 24-well plates were transfected with plasmids using Lipofectamine 2000 (Invitrogen) according to the manufacturer's instructions. At 48 h post transfection, the cells were fixed with methanol, stained with Wright-Giemsa staining, and examined with light microscope (Leica, Wetzlar, Germany). Alternatively, cells transfected with the plasmid pEGFP-NS22 were fixed with 4% paraformaldehyde at 48 h post transfection. The fixed cells were stained with Hoechst 33342 and observed by fluorescence microscopy as described previously [38,39].

Phylogenetic Analysis
Phylogenetic analysis was performed based on the alignment of the concatenated sequences of seven structural proteins that are conserved in all sequenced aquareoviruses ( Table 1). The seven structural proteins from 12 other aquareoviruses were rearranged as continuous amino acid sequences with the same order as MsReV. The concatenated protein sequences were then aligned with the Clustal X 1.83 program, and phylogenetic tree was constructed using the neighbor-joining method with 1000 bootstrap replicates in MEGA5 software [40]. GenBank accession numbers of the aquareovirus sequences used for analysis were shown in Table 1.

Identification of MsReV and Electrophoretic Pattern of Viral Genome Segments
The filtered tissue homogenates from diseased largemouth bass caused overt cytopathic effects (CPE) in GCF, GCO, CHSE-214 and BF-2 cell lines after three to four days incubation. The temperature range for virus replication extended from 15-25˝C, and the maximum virus infectivity titer was obtained in GCF cell lines with 1ˆ10 5.5 TCID 50 mL´1 at 20˝C. Electron microscopic observations of negatively-stained samples revealed that the virus particles have the typical morphology of aquareoviruses, which have characteristic double-layered capsids and are approximately 70-80 nm in diameter (Figure 1a). This virus is now referred to as largemouth bass Micropterus salmoides reovirus (MsReV).

Identification of MsReV and Electrophoretic Pattern of Viral Genome Segments
The filtered tissue homogenates from diseased largemouth bass caused overt cytopathic effects (CPE) in GCF, GCO, CHSE-214 and BF-2 cell lines after three to four days incubation. The temperature range for virus replication extended from 15-25 °C, and the maximum virus infectivity titer was obtained in GCF cell lines with 1 × 10 5.5 TCID50 mL −1 at 20 °C. Electron microscopic observations of negatively-stained samples revealed that the virus particles have the typical morphology of aquareoviruses, which have characteristic double-layered capsids and are approximately 70-80 nm in diameter (Figure 1a). This virus is now referred to as largemouth bass Micropterus salmoides reovirus (MsReV).  MsReV genomic dsRNAs were extracted from purified virus particles and analyzed by polyacrylamide gel electrophoresis (PAGE). The genomic dsRNAs from two other aquareoviruses, SMReV and GCReV-109, as previously described by our laboratory [24,26], were also prepared and used as molecular weight size markers. The GCReV-109 genome segments have been divided into four size clusters: cluster 1 (S1-S3), cluster 2 (S4-S6), cluster 3 (S7-S9) and cluster 4 (S10 and S11), and the migration pattern of GCReV-109 genome segments described as 3-3-3-2 is typical for grass carp group (freshwater fish) aquareovirus. SMReV genome segments were separated into 10 distinct bands, with segments S1 and S2 comigrating, and the migration pattern of SMReV genome segments described as 3-3-2-2-1, is the other group (saltwater fish) of aquareovirus. Here, the electrophoretic migration pattern of MsReV genomic dsRNA in polyacrylamide gel was shown in Figure 1b (left panel). The genome segments of MsReV were separated into nine distinct bands, with segments S1 and S2, and segments S7 and S8 comigrating, respectively. The migration pattern of MsReV genome segments was clearly identified as 3-3-2-2-1. The characteristics of genome segments and predicted proteins of MsReV were shown in Figure 1b (right panel). The results showed that the electropherotype of MsReV genome segments was similar to that of SMReV, but some changes were apparent in GCReV-109 genome electropherotype.

Organization and Structure of MsReV Genome
The genome segments 1-11 of MsReV were sequenced completely and have been deposited in GenBank under accession numbers KJ740724 to KJ740734. The complete genome sequence of MsReV consists of 24,024 bp divided into 11 segments that range in size from 3947 bp (S1) to 783 bp (S11) (Figure 1b, right panel). The lengths of non-coding regions (NCRs) of MsReV genome segments ranged from 12-28 bp at the 5 1 ends, and ranged from 29-162 bp at the 3 1 ends. Analysis of the 5 1 -and 3 1 -NCRs showed that all of MsReV segments shared conserved nucleotides with 5 1 -GUUUUA U /G/ A at their 5 1 ends and U / A UUCAUC-3 1 at their 3 1 ends. Moreover, the first and last nucleotides of all segments were complementary (G-C), which are known to be highly conserved within aquareoviruses.
The open reading frame (ORF) analysis revealed that all of MsReV genome segments contained a single ORF, with the exception of S7 segment, which had two partially overlapping ORFs. MsReV was predicted to encode a total of 12 proteins, including seven structural proteins (VP1 to VP7) and five nonstructural proteins (NS87, NS22, NS32, NS38 and NS25) (Figure 1b, right panel). Alignments of the concatenated sequences of seven structural proteins from 13 aquareoviruses revealed that MsReV showed high sequence identities with SMReV (91.2%) and chum salmon reovirus (CHSRV) (85.5%) from hosts in saline environments, but a low level of identity (30.4%-48.5%) with other aquareoviruses from hosts in freshwater environments (Table 1).

Comparison of the S4 Segments and Encoded Nonstructural Proteins of Three Aquareoviruses
Nucleic acid sequence analysis indicated that the S4 genome segments of MsReV, SMReV and GCReV-109 were predicted to encode three homologous nonstructural proteins (NS87, NS88 and NS79, respectively), which were thought to be involved in the formation of viral inclusion bodies. MsReV NS87 was about 87 kDa, consisting of 811 amino acids, which was a little smaller than SMReV NS88 (88 kDa, 817 aa), but larger than GCReV-109 NS79 (716 aa, 79 kDa). Amino acid sequence alignment showed that MsReV NS87 shared 69.5% identity to SMReV NS88, and only 16.1% identity to GCReV-109 NS79. As predicted by Coils program, MsReV NS87 had two coiled coils (Coil 1, aa 582-632, and Coil 2, aa 688-756) and an intercoil spacing between the two coils, which were also conserved in SMReV NS88 ( Figure 2). However, only one coil, corresponding to the regions of Coil 2 in MsReV NS87 and SMReV NS88, was predicted at amino acid positions 582-640 of GCReV-109 NS79 (Figure 2).

Figure 2.
Diagram of predicted coil regions in the nonstructural proteins encoded by S4 segments among MsReV, SMReV and GCReV-109. MsReV NS87 and SMReV NS88 have two coiled coils (Coil 1 and Coil 2) and an intercoil spacing between the two coils, while GCReV-109 NS79 has only one coil (Coil 2). Numbers refer to amino acids residues.

Experimental Verification of the Function of NS22 Protein Encoded by MsReV S7
Initial infection assay showed that GCF cell line was susceptible to MsReV, and the infected cells displayed typical CPE characterized by cell-cell fusion and syncytium formation (Figure 3a). Subsequently, a fusion-associated small transmembrane (FAST) protein NS22 was identified in MsReV S7 segment by sequence analysis. The FAST protein is a viral nonstructural protein, which has membrane-destabilizing activity that may contribute to cell-cell fusion and syncytium formation in virus-infected cells. To confirm the ability of MsReV NS22 to induce cell-cell fusion, we generated series of recombinant plasmids which contain different regions of S7, including the full gene encoding NS22 with enhanced green fluorescent protein (pEGFP-NS22) or without EGFP (1-613), deletions (14-613, 15-613, and 17-613) and point mutations (1-613/∆14 and 1-613/∆18) for protein expression in GCF cells. Expression of the full gene (NS22-EGFP and 1-613), deletion (14-613) and point mutation (1-613/∆14) induced multinucleated syncytia formation in GCF cells (Figure 3b,c), but the deletions (15-613) and (17-613) did not have any noticeable effect on the phenotype in the cells. Furthermore, for detection of NS22 activity, the start codon for NS22 ORF was disrupted by changing a single nucleotide at 18 ( 17 CUG 19 to CCG). The gene expression that alteration of the translational start site prevents synthesis of MsReV NS22 and syncytium formation in transfected cells were assessed, point mutation (1-613/∆18) could not produce syncytia (Figure 3c). The experiments showed that MsReV NS22 encoded by S7, which is translated from a CUG start codon, indeed, contributed to syncytial cytopathic effect. The S7-coded proteins (e.g., NS22, NS32) of MsReV and other aquareoviruses were depicted and compared. MsReV NS22 showed high similarity in structure and function of SMReV NS22 previously reported by our laboratory [26], and they were clustered in one subgroup (host in saline environments). The small size NS16 proteins were clustered in another subgroup, including American grass carp

Experimental Verification of the Function of NS22 Protein Encoded by MsReV S7
Initial infection assay showed that GCF cell line was susceptible to MsReV, and the infected cells displayed typical CPE characterized by cell-cell fusion and syncytium formation (Figure 3a). Subsequently, a fusion-associated small transmembrane (FAST) protein NS22 was identified in MsReV S7 segment by sequence analysis. The FAST protein is a viral nonstructural protein, which has membrane-destabilizing activity that may contribute to cell-cell fusion and syncytium formation in virus-infected cells. To confirm the ability of MsReV NS22 to induce cell-cell fusion, we generated series of recombinant plasmids which contain different regions of S7, including the full gene encoding NS22 with enhanced green fluorescent protein (pEGFP-NS22) or without EGFP (1-613), deletions (14-613, 15-613, and 17-613) and point mutations (1-613/∆14 and 1-613/∆18) for protein expression in GCF cells. Expression of the full gene (NS22-EGFP and 1-613), deletion (14-613) and point mutation (1-613/∆14) induced multinucleated syncytia formation in GCF cells (Figure 3b,c), but the deletions  and (17-613) did not have any noticeable effect on the phenotype in the cells. Furthermore, for detection of NS22 activity, the start codon for NS22 ORF was disrupted by changing a single nucleotide at 18 ( 17 CUG 19 to CCG). The gene expression that alteration of the translational start site prevents synthesis of MsReV NS22 and syncytium formation in transfected cells were assessed, point mutation (1-613/∆18) could not produce syncytia (Figure 3c). The experiments showed that MsReV NS22 encoded by S7, which is translated from a CUG start codon, indeed, contributed to syncytial cytopathic effect. The S7-coded proteins (e.g., NS22, NS32) of MsReV and other aquareoviruses were depicted and compared. MsReV NS22 showed high similarity in structure and function of SMReV NS22 previously reported by our laboratory [26], and they were clustered in one subgroup (host in saline environments). The small size NS16 proteins were clustered in another subgroup, including American grass carp reovirus (AGCRV), golden shiner reovirus (GSRV) and GCRV-873 (host in freshwater environments), or even lacked the corresponding gene from GCReV-109 (host in freshwater environment) (Figure 3d).

MsReV and SMReV Are Closely Related to Equivalent Genome Segments
Functionally equivalent genome segments from the three aquareoviruses, MsReV, SMReV and GCReV-109 were analyzed. The MsReV and SMReV genome segments encode 12 proteins, respectively, which consist of seven structural proteins (VP1 to VP7) and five nonstructural proteins (NS87 or NS88, NS22, NS32, NS38 and NS25). However, GCReV-109 genome segments encode only 11 proteins, which consist of seven structural proteins and four nonstructural proteins (NS79, NS56, pun and NS38). GCReV-109 lacks the genes encoding the nonstructural proteins (NS22 and NS25), and S8 encodes a protein of unknown function (pun) that has no equivalent protein in MsReV and SMReV. These results showed that MsReV was more closely related to SMReV than to GCReV-109 ( Figure 4).

MsReV is More Closely Related to SMReV than to GCRV-109
The analysis above revealed that MsReV and SMReV shared a close relationship through genome anatomy and gene function detection but the correlation between their host environments is poorly resolved. To check the association between aquareoviruses' phylogenetic background and their host environments, we undertook to further characterize and compare genome segments encoding multiple 3.6. MsReV Is More Closely Related to SMReV than to GCRV-109 The analysis above revealed that MsReV and SMReV shared a close relationship through genome anatomy and gene function detection but the correlation between their host environments is poorly resolved. To check the association between aquareoviruses' phylogenetic background and their host environments, we undertook to further characterize and compare genome segments encoding multiple proteins. Thirteen aquareoviruses were analyzed according to the concatenated sequences of seven structural proteins (VP1 to VP7) ( Table 1). The phylogenetic tree showed that the 13 aquareoviruses were divided into two subgroups, one is host in freshwater environments and the other is host in saline environments ( Figure 5). MsReV was closely clustered with SMReV in the subgroup of host in saline environments, and GCRV-109 was clustered with AGCRV, GSRV and different GCRV isolates in the subgroup of host in freshwater environments. The evidence that aquareoviruses species closely related also have similar host environments indicated that the high genomic structure similarities between MsReV and SMReV were associated with their hosts in saline environments.
Viruses 2015, 7 16 proteins. Thirteen aquareoviruses were analyzed according to the concatenated sequences of seven structural proteins (VP1 to VP7) ( Table 1). The phylogenetic tree showed that the 13 aquareoviruses were divided into two subgroups, one is host in freshwater environments and the other is host in saline environments ( Figure 5). MsReV was closely clustered with SMReV in the subgroup of host in saline environments, and GCRV-109 was clustered with AGCRV, GSRV and different GCRV isolates in the subgroup of host in freshwater environments. The evidence that aquareoviruses species closely related also have similar host environments indicated that the high genomic structure similarities between MsReV and SMReV were associated with their hosts in saline environments.

Discussion
Present research involves whole-genome sequencing and electrophoretic migration pattern of MsReV genome segments, functional identification of proteins, and phylogenetic analysis of concatenated structural protein sequences. These studies revealed marked similarities between the genomic structures of a novel aquareovirus MsReV and SMReV. These aquareoviruses from hosts that had similar environments (e.g., saline environments) were closely related, but alienated from those whose hosts lived in different environments (e.g., freshwater environments). These findings suggest that the aquareoviruses' genome diversity is associated with their host environments. For example, MsReV,

Discussion
Present research involves whole-genome sequencing and electrophoretic migration pattern of MsReV genome segments, functional identification of proteins, and phylogenetic analysis of concatenated structural protein sequences. These studies revealed marked similarities between the genomic structures of a novel aquareovirus MsReV and SMReV. These aquareoviruses from hosts that had similar environments (e.g., saline environments) were closely related, but alienated from those whose hosts lived in different environments (e.g., freshwater environments). These findings suggest that the aquareoviruses' genome diversity is associated with their host environments. For example, MsReV, SMReV and CHSRV were isolated from different locations and at different times [26,41], but they were clustered into the same clade in the phylogenetic tree. Thus, the high genomic structure similarities among the three aquareoviruses might have no relation to the space and time of virus isolation, but were related to the physiological environment of their hosts. This is the first report indicating that aquareoviruses' genomic structures are associated with their host physiological conditions. Broadening our understanding of the genomic diversity of aquareoviruses that exist in different host environments will significantly improve our ability to recognize novel aquareoviruses in the context of aquaculture disease outbreaks.
The fusion-associated small transmembrane (FAST) proteins of the fusogenic reoviruses are the only known examples of membrane fusion proteins encoded by nonenveloped viruses [26,[42][43][44]. MsReV S7 encodes the FAST protein NS22 with a CUG start codon, which contributes to syncytial cytopathic effect. The results were confirmed by the experiments of construction and expression of plasmids carrying different regions of S7, and might be further expanded to study not only the response of the MsReV FAST protein function, but also that aquareoviruses were divided into two major branches of host environments according to the FAST protein structure (except for CHSRV). More significantly, investigating the correlation of aquareovirus genetic variants with viral host environments through the FAST proteins could provide the first link between aquareovirus nonstructural protein, gene structure and its host environment, allowing us to establish the relevance and causal relationship of aquareovirus function genes to their host environments.
An increasing number of different reoviruses have been isolated from marine and fresh water in the past years [8,28,34,45]. Emergence of new infectious diseases in lower vertebrates (e.g., fish) or human is not a new phenomenon [46]. Viruses are very genetically diverse and new genotypes, strains and species evolve rapidly [47]. It is important to increase understanding of the virus genomic structure, and several factors are believed to be major reasons for generating genetic variation in RNA viruses, such as mutation [48,49], recombination [50], and reassortment [51,52]. Aquareoviruses' genomic diversity provides a unique opportunity to examine and explore the mechanisms that are involved in the evolution of multisegmented RNA virus genome and their host environments [4]. Comparative genomic sequencing, functional characterization of two nonstructural proteins combined with functionally equivalent genome segments analysis, and in particular the phylogenetic analysis of the concatenated sequences of seven structural proteins indicated that aquareoviruses' extensive genomic structural variations between hosts in saline environments and freshwater environments, and host environments similar to aquareoviruses, occur most often between closely related species in natural populations. Although these studies have offered new insights into the host environment's effect on aquareovirus genomic structure, the mechanisms involved in viral genomic structural variations impacted by host environments need to be further explored.

Author Contributions
Q.Y.Z. conceived and designed the experiments; Z.Y.C. and X.C.G. performed the experiments; Q.Y.Z. and Z.Y.C. analyzed the data and wrote the paper. All authors read and approved the final manuscript.

Conflicts of Interest
The authors declare no conflict of interest.