From the assembled virus clones, the full length cDNA sequences of three dsRNA segments were obtained. Subsequent sequence analysis indicated that the 6-kbp large segment was the genome of a novel fusarivirus that we named NsFV1 (Accession number: MT774523), while the other two dsRNA segments, in approximately 2-kbp, constituted the genome of a novel partitivirus, NsPV1 (Accession number: MT774524 and MT774525). Schematic diagram of genomic organization of the two viruses were represented in Fig. 1A and Fig. 2A,respectively.
The complete genome of NsFV1 was 6,147 bp (excluding the polyA tail), with a G + C content of 52.4%. It contained two ORFs (ORF1 and 2) on the genomic RNA, preceded and followed by untranslated regions (UTRs) of 77 nt and 53 nt, respectively. NsFV1 ORF1 encoded a 1,508 amino-acid (aa) polyprotein with a calculated molecular mass of 169.2 kDa. Conserved domain search against the NCBI database revealed the presence of an conserved RdRp (RdRP_1, pfam00680) and a helicase (DEAD, pfam00270) domains in the ORF1-encoded polyprotein. Eight conserved motifs were found in the conserved RdRp domain by multiple aa sequence alignment and comparison between the viral RdRps between NsFV1 and other selected viruses (Fig. 1B). Homology search, using the BLASTp, indicated that the ORF1-encoded protein of NsFV1 had significant sequence identities to the proteins encoded by other putative fusariviruses, such as Gaeumannomyces tritici fusarivirus 1 (GtFV1), Plasmopara viticola associated fusarivirus 2 (PvAFV2), Penicillium aurantiogriseum fusarivirus 1 (PaFV1). The RdRp domain of NsFV1 showed 38.87–63.83% aa sequence identities to the corresponding domains of other fusariviruses, with GtFV1 being the best math displaying 63.83% aa identity (Query coverage: 99%; E value: 1e-142). In addition, the RdRp domain of NsFV1 also showed moderate levels of sequence identity (Identity: 25.55%-28.43%; Query coverage: 64%-92%) to that of the hypoviruses in the family Hypoviridae, and that of other related viruses. Similarly, the Hel domain of NsFV1 had a significant degree of aa sequence identity to those of the fusariviruses, ranging from 29.77–58.09% (Query coverage: 73 to 100%). However, the NsFV1 Hel domain had no sequence identity to those of the viruses in the family Hypoviridae, but had low sequence identity to some bacterial RNA helicases. NsFV1 ORF2 was predicted to encode a 45.8 kDa protein with 411 aa. Database search revealed that the ORF2-encoded protein was homologous only to the hypothetical protein encoded by GtFV1, and showed no significant similarity to any other known viral proteins. However, a putative conserved domain of SMC (structural maintenance of chromosomes, COG1196) was found in the N-half of the ORF2 encoded protein, which was predicted to be associated with cell cycle control, cell division and chromosome partitioning. The SMC proteins were found in bacteria, archaea and eukaryotes. However, some viruses including fusariviruses also contained the SMC domain as described previously . Its occurrence in mycovirus is noteworthy and might reveal the horizontal gene transfer between viruses and other eukaryotes or prokaryotes. In the other hand, the presence of SMC in different viruses might also indicate the common origin of these viruses. The true function and evolution of the SMC domain in the NsFV1 ORF2 needed further study.
In order to estimate the phylogenetic relationships between NsFV1 and other mycoviruses, we conducted phylogenetic analysis based on aa alignments of the RdRp and Hel domains. The phylogenetic tree generated based on the RdRp domain alignment showed that NsFV1 was clustered in a fusarivirus-clade, including the prototype FgV1, and distinct from members of the family Hypoviridae (Fig. 1C). A phylogenetic tree of the Hel domain also showed that NsFV1 was grouped in a clade and exhibited a similar tree topology with the RdRp based phylogenetic analysis (Fig. S1). Overall, based on the similarities in conserved RdRp and Hel domains and phylogenetic analysis, we can suggest the NsFV1 as a novel member of the proposed family Fusariviridae.
Sequences analysis indicated that the NsPV1 was composed of two dsRNA segments, which were designated as dsRNA 1 and dsRNA 2, respectively. NsPV1 dsRNA 1 was 1,796 bp in length, with a G + C content of 47.6%, while the dsRNA 2 was 1,455 bp long, containing a 51.2% G + C content. The 5′ and 3′ untranslated regions (UTRs) were 104 bp and 75 bp in dsRNA 1, 92 bp and 100 bp in dsRNA 2, respectively. The 5’ and 3’UTRs between the dsRNA 1 and dsRNA 2 were conserved, possessing an identical stretch of 5’-CGTGAAATAC-3’ in the 5’-terminal and a stretch of 5’-TAAACCAAAA-3’ in the 3’-terminal (Fig. 2B), which were considered to be important for replication of viruses with multicomponent RNA genome .
The coding strand of the NsPV1 dsRNA 1 contained a single ORF encoding a 538 aa protein with an estimated molecular mass of 62.0 kDa. Homology search revealed that the dsRNA 1 encoded protein shared sequence similarity with RdRp proteins encoded by other partitiviruses. Magnaporthe oryzae partitivirus 1 (MoPV1) was the most closely related virus, showing aa sequence identity of 70.50% (E value: 0; query cover: 100%), followed by Magnaporthe oryzae partitivirus 2 (MoPV2, identity: 68.65%; E value: 0; query cover: 100%), Penicillium stoloniferum virus F (PsV-F, (identity: 67.23%; E value: 0; query cover: 99%) and Pythium nunn virus 1 (PnV1, identity: 55.51%; E value: 0; query cover: 95%). In addition, conserved motifs characteristic of RdRp sequences of other members of the Partitiviridae were found in the NsPV1 encoded RdRp. NsPV1 dsRNA 2 contained a single ORF, which was predicted to encode a putative 46.5 kDa protein comprised of 420 aa. The 46.5 kDa protein shared the maximal aa identity of 65.53% (E value: 0; query cover: 96%) to that of MoPV1, followed by MoPV2, PsV-F and PnV1, with the aa identities ranging from 37.50–65.28%. Interesting, the 46.5 kDa protein also showed a lower level of aa sequence identity to proteins of other eukaryote, including Gracilariopsis chorda, Oidium neolycopersici and Erysiphe necator. A phylogenetic tree was constructed using the RdRp of NsPV1 and other partitiviruses (Fig. 2C). It revealed that NsPV1 was placed in a branch including members of the genus Gammapartitivirus, as expected from the homology search result.
The species demarcation criteria of partitiviruses are that aa identities in the RdRp and CP is fewer than 90% and 80%, respectively . Since the aa identities of RdRp and CP between NsPV1 and the closely related MoPV1 were lower than the cutoff values, we can suggest the classification of NsPV1 as a novel species of the genus Gammapartitivirus in the family Partitiviridae.
In conclusion, we identified two novel mycoviruses from a N. sphaerica strain Ns2-3. Sequence similarity, genome organization and phylogenetic analysis supported the affiliation of NsFV1 and NsPV1 as novel members of the family Fusariviridae and Partitiviridae, respectively. As far as we know, this was the first report of mycoviruses co-infected in the fungus N. sphaerica. The biological effects of these viruses on the fungal host and the functions of the virus genes in terms of virus-host interactions remained further study.