Diversity of Mycoviruses Present in Strains of Binucleate Rhizoctonia and Multinucleate Rhizoctonia, Causal Agents for Potato Stem Canker or Black Scurf

In this study, the diversity of putative mycoviruses present in 66 strains of binucleate Rhizoctonia (BNR, including anastomosis group (AG)-A, AG-Fa, AG-K, and AG-W) and 192 strains of multinucleate Rhizoctonia (MNR, including AG-1-IA, AG-2-1, AG-3 PT, AG-4HGI, AG-4HGII, AG-4HGIII, and AG-5), which are the causal agents of potato stem canker or black scurf, was studied using metatranscriptome sequencing. The number of contigs related to mycoviruses identified from BNR and MNR was 173 and 485, respectively. On average, each strain of BNR accommodated 2.62 putative mycoviruses, while each strain of MNR accommodated 2.53 putative mycoviruses. Putative mycoviruses detected in both BNR and MNR contained positive single-stranded RNA (+ssRNA), double-stranded RNA (dsRNA), and negative single-stranded RNA (-ssRNA) genomes, with +ssRNA genome being the prevalent nucleic acid type (82.08% in BNR and 75.46% in MNR). Except for 3 unclassified, 170 putative mycoviruses found in BNR belonged to 13 families; excluding 33 unclassified, 452 putative mycoviruses found in MNR belonged to 19 families. Through genome organization, multiple alignments, and phylogenetic analyses, 4 new parititviruses, 39 novel mitoviruses, and 4 new hypoviruses with nearly whole genome were detected in the 258 strains of BNR and MNR.


Extraction of Total RNA
For extracting total RNA, the 258 strains of BNR and MNR were cultured on PDA plates covered with cellophane film membranes (PDA-CF) at 25 °C in the dark for five days. Approximately 0.5 g of fresh mycelia were harvested from PDA-CF and ground to fine powder in liquid nitrogen, and then total RNA was extracted using TRIzol Reagent (Invitrogen, Carlsbad, CA, USA) according to the manufacturer's protocol. The concentrations and quality of RNA samples were measured using an ultramicro spectrophotometer (Nanodrop 2000, Thermo Scientific, Waltham, MA, USA). To confirm the RNA integrity, 1.0% (w/v) gel electrophoresis agarose was used. Finally, RNA samples were pooled to have the same final concentration (~200 ng/µL), resulting in two pools which were from BNR and MNR, respectively. . Geographic origins where Rhizoctonia isolates were collected. The size of the red spots represented the number of Rhizoctonia isolates collected from each province, municipality, or autonomous region. Abbreviation and full name of provinces, municipalities, and autonomous regions are as follows: AH, Anhui province; BJ, Beijing municipality; CQ, Chongqing municipality; FJ, Fujian province; GD, Guangdong province; GS, Gansu province; GX, Guangxi Zhuang autonomous region; GZ, Guizhou province; HeB, Hebei province; HeN, Henan province; HLJ, Heilongjiang province; HuB, Hubei province; IM, Inner Mongolia autonomous region; JL, Jilin province; JS, Jiangsu province; LN, Liaoning province; NX, Ningxia Hui autonomous region; QH, Qinghai province; SC, Sichuan province; SN, Shaanxi province; SX, Shanxi province; XJ, Xinjiang Uygur autonomous region; YN, Yunnan province; ZJ, Zhejiang province.

Extraction of Total RNA
For extracting total RNA, the 258 strains of BNR and MNR were cultured on PDA plates covered with cellophane film membranes (PDA-CF) at 25 • C in the dark for five days. Approximately 0.5 g of fresh mycelia were harvested from PDA-CF and ground to fine powder in liquid nitrogen, and then total RNA was extracted using TRIzol Reagent (Invitrogen, Carlsbad, CA, USA) according to the manufacturer's protocol. The concentrations and quality of RNA samples were measured using an ultramicro spectrophotometer (Nanodrop 2000, Thermo Scientific, Waltham, MA, USA). To confirm the RNA integrity, 1.0% (w/v) gel electrophoresis agarose was used. Finally, RNA samples were pooled to have the same final concentration (~200 ng/µL), resulting in two pools which were from BNR and MNR, respectively.

Metatranscriptome Sequencing
Metatranscriptome sequencing of the 258 strains of BNR and MNR was conducted by Shanghai Biotechnology Corporation using an Illumina X-TEN instrument with pairedend program. TruSeq Stranded Total RNA LT Sample Prep Kit (Illumina, San Diego, CA, USA) was used to establish sequencing libraries of strains of BNR and MNR from rRNA-depleted total RNA. Library quality was checked using Qubit ® 2.0 Fluorometer (Invitrogen, Q32866) and Agilent Technologies 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA, USA). The low-quality reads were filtered out to obtain high-quality clean reads. De novo sequence assembly was constructed using CLC Genomics Workbench version 6.0.4 software. The resulting final sequences were subjected to National Center for the Biotechnology Information (NCBI) non-redundant (NR) database and aligned using BLASTx to confirm the mycovirus-like contigs and classification status, nucleic acid type, and the virus best matched of these contigs.

Genome Organization and Phylogenetic Analysis
NCBI ORF finder program (https://www.ncbi.nlm.nih.gov/orffinder/, accessed on 1 June 2022) was used to predict open reading frames (ORFs) of contigs related to mycoviruses obtained from metatranscriptome sequencing based on standard genetic code or fungal mitochondrial genetic code (whose number is 4). BLASTp and BLASTx were used to search for homologous mycoviruses against NCBI NR database. When the e-value is less than or equal to 1 × 10 −5 , the annotation result is considered reliable. Conserved Domain Database (CDD) (http://www.ncbi.nlm.nih.gov/Structure/cdd/wrpsb.cgi, accessed on 1 June 2022), Protein Family (Pfam) database (http://pfam.sanger.ac.uk/, accessed on 1 June 2022), and PROSITE database (http://www.expasy.ch/, accessed on 1 June 2022) were used to find conserved motifs of amino acid (aa) sequences of putative mycoviruses. The aa sequences alignments were conducted by CLUSTAL_X [54]. Phylogenetic trees based on aa sequences of RNA dependent RNA polymerase (RdRp), or polyprotein were conducted by the maximum likelihood (ML) method in Jones-Taylor-Thornton (JTT) model with 1000 bootstrap replicates using MEGA software version 6.0 [55].

Virus Names
The name of a novel putative mycovirus identified in this study for the first time is named according to a previous reference [56], which consists of three parts: (I) the first part of the name is the source of the virus; (II) the second part of the name shows the virus taxonomical group; and (III) the third part of the name is a progressive number [57]. For example, "Rhizoctonia solani [part I] partitivirus [part II] 12 [part III]" presents a new partitivirus and the twelfth partitivirus found in R. solani. A mycovirus previously reported, which was also identified in this study, was labeled with "BNR" or "MNR" to indicate its sources in this study. For example, "Rhizoctonia solani partitivirus 6-BNR" presents a strain of Rhizoctonia solani partitivirus 6 reported previously [58] and is identified from a strain of BNR in this study.

Comparison of Mycoviral Diversity in BNR and MNR
Metatranscriptome sequencing was conducted based on the libraries of strains of BNR and MNR, and 11.34 GB raw data and 11.22 GB raw data were obtained from the libraries of strains of BNR and MNR, respectively. After filtering and de novo assembly, 99,951 contigs and 112,939 contigs were acquired from strains of BNR and MNR, respectively ( Table 1). Homology of the contigs with more than 200 nt in length were conducted by screening against GenBank NR database, the most detailed protein database currently used for protein function and structure annotation. The best matches of these contigs were determined based on the top hit (the highest identity and query cover to the available viral genomes in the NR database) from BLASTx and e-values (≤1 × 10 −5 ). As a result, a total of 658 contigs were best matched with viral genomes, with 173 contigs and 485 contigs being found in BNR and MNR, respectively. Annotation results showed that 472 (71.7%) out of these 658 contigs had low identities (≤70%) of aa sequences with mycoviruses reported previously. Contig ids, length, best match (most closely related viruses), query cover, amino acid identity, and e-value were listed in Table S2. The results of BLASTx analysis showed that three and 33 contigs related to viruses obtained from BNR and MNR, respectively, were best matched with unclassified viruses reported previously and considered as putative unclassified mycoviruses. For example, First_Contig8786 with 1736 nt in length was most closely related to an unclassified virus, soybean leaf-associated negative-stranded RNA virus 4. The remaining 170 putative mycoviruses found in BNR and the remaining 452 putative mycoviruses found in MNR belonged to 13 and 19 virus families, respectively. The 11 virus families found in both BNR and MNR were Benyviridae, Botourmiaviridae, Endornaviridae, Fusariviridae, Hypoviridae, Mitoviridae, Narnaviridae, Partitiviridae, Totiviridae, Mymonaviridae, and a proposed family Bipartitiviridae. Moreover, the two families (Virgaviridae and Rhabdoviridae) were only detected in BNR, while the eight families (Betaflexiviridae, Bunyaviridae, Deltaflexiviridae, Gammaflexiviridae, Megabirnaviridae, Togaviridae, Tombusviridae, and Tymoviridae) were only discovered in MNR (Table 2).
Negative single-stranded RNA (-ssRNA) Total 173 485 Note: " a " indicated that no virus belonging to the eight families Betaflexiviridae, Deltaflexiviridae, Gammaflexiviridae, Tymoviridae, Togaviridae, Tombusviridae, Megabirnaviridae, and Bunyaviridae was detected in BNR; " b " indicated that no virus belonging to the two families Virgaviridae and Rhabdoviridae was detected in MNR.

Genome Organization and Phylogenetic Analysis of Putative Members of the Family Mitoviridae
The 78 and 149 contigs related to mycoviruses associated with the family Mitoviridae were found in BNR and MNR, respectively. Among the 227 contigs, 54 contigs (26 contigs found in BNR and 28 contigs found in MNR) whose lengths are longer than 3000 nt or which can encode complete RdRp was chosen to predict open reading frame (ORF) and perform multiple alignments. The contig ids, best matches, names, and GenBank accession numbers of the 54 mitoviruses were listed in Table 4. The results showed that among the 54 contigs, the lengths of aa sequences of 50 contigs (92.59%) were over 700 aa, and the lengths of aa sequences of 21 contigs (38.89%) were over 900 aa. Since the lengths of complete aa sequence of RdRp of most mitoviruses submitted to the NCBI NR database (https://www.ncbi.nlm.nih.gov/protein, accessed on 1 June 2022) range from 500 aa to 900 aa, the lengths of aa sequences of RdRp of these 54 mitoviruses in this study are relatively longer than that of most mitoviruses reported previously.
Among these fifty-four contigs, the length of aa sequence of four contigs (one from BNR and three from MNR) was shorter than 700 aa, the length of aa sequence of nine contigs (seven from BNR and two from MNR) ranged from 700 aa to 800 aa, the length of aa sequence of twenty contigs (eight from BNR and 12 from MNR) ranged from 800 aa to 900 aa, the length of aa sequence of twelve contigs (six from BNR and six from MNR) ranged from 900 aa to 1000 aa, and the length of aa sequence of nine contigs (four from BNR and five from MNR) were longer than 1000 aa. Ten contigs with these five length ranges (five contigs from BNR and five contigs from MNR) were chosen (Table 4) and used for genome organization and multiple alignments (Figure 3). All the ten mitoviruses contained an ORF encoding RdRp ( Figure 3A) based on mitochondrial codon usage. Through multiple alignments of aa sequences of RdRp of these ten mitoviruses and their homologous viruses, four conserved motifs (motif III-VIII) were found ( Figure 3B) in the viral RdRp domain; moreover, GDD tripeptide was found in motif IV ( Figure 3B).  Note: " a " indicated novel mitoviruses found in BNR or MNR; " b " indicated mycoviruses selected for genome organization and multiple alignment analyses.  Phylogenetic analysis based on the aa sequences of RdRp showed that these 54 putative mycoviruses with nearly the whole genome found in BNR and MNR clustered into the family Mitoviridae. Except for contig2511 (Rhizoctonia solani mitovirus 1-BNR) which was clustered in clade I ( Figure 3C), the remaining 53 contigs clustered together in clade II. Red fonts and green fonts indicate the mitoviruses found in BNR and MNR, respectively. A total of 38 (18 found in BNR and 20 found in MNR) out of the 54 mitoviruses discovered in this study are new species of the family Mitoviridae (Table 4).

Genome Organization and Phylogenetic Analysis of Putative Members of the Family Hypoviridae
One contig found in BNR and three contigs found in MNR with >8000 nt in length or encoding complete ORF were most closely related to hypoviruses, which were named Binucleate Rhizoctonia hypovirus 1 (BRHV1), Rhizoctonia solani hypovirus 9 (RsHV9), Rhizoctonia solani hypovirus 10 (RsHV10), and Rhizoctonia solani hypovirus 11 (RsHV11), respectively. The contig ids, best matches, names, and GenBank accession numbers of the four hypoviruses were listed in Table 5. Analysis of genome organization showed that these four hypoviruses all contained one +ssRNA encoding polyprotein ( Figure 4A). The polyprotein of RsHV9 contained RdRp and RNA helicase (Hel) domains. The polyproteins of RsHV10 and BRHV1 contained one and two Hel domains, respectively. The polyprotein of RsHV11 contained DUF4286 and Hel domains. The Hel domain was found in the polyprotein of all the four hypoviruses, while the RdRp, necessary for mycovirus replication, was only contained in polyprotein of RsHV9. The similar characteristic had been documented in other members of the family Hypoviridae; for example, the RdRp domain was contained in polyprotein of Fusarium graminearum hypovirus 2 (FgHV2) [21] but was not contained in polyprotein of Alternaria alternata hypovirus 1 (AaHV1) [62]. Multiple alignment of aa sequences of Hel of the four hypoviruses and their homologous viruses revealed that three conserved motifs (motif I-III) ( Figure 4B) were found in the Hel domain.   Phylogenetic analysis based on the aa sequences of polyprotein of the four hypoviruses and their homologous viruses showed that RsHV10 clustered into the genus Gammahypovirus, and RsHV9 and RsHV11 clustered together with unclassified hypoviruses, such as Sclerotium rolfsii hypovirus 3, 4, and 7 [63]. BRHV1 clustered into the family Hypoviridae, but BRHV1 was relatively distinct from other hypoviruses within the family Hypoviridae ( Figure 4C).

Discussion
Our study represents in detail the first record of the putative mycoviruses associated with strains of BNR and MNR causing potato stem canker or black scurf in China and perhaps worldwide using metatranscriptome sequencing. In this study, putative mycoviruses belonging to eight families (Benyviridae, Botourmiaviridae, Fusariviridae, Hypoviridae, Mymonaviridae, Rhabdoviridae, Totiviridae, and Virgaviridae) and seven families (Betaflexiviridae, Botourmiaviridae, Deltaflexiviridae, Gammaflexiviridae, Mymonaviridae, Togaviridae, and Tombusviridae) were firstly found in BNR and MNR, respectively.
Metatranscriptome sequencing is widely used to discover mycoviruses in different host species and has supported the progress of research in virus pathogenesis and controlling of related diseases [33]; however, the lengths of some contigs related to mycoviruses obtained from metatranscriptome sequencing were less than 1000 nt, some of which might only cover 1% of the whole genome of the corresponding mycoviruses. In this study, there were 60 endornaviruses (84.51%) with less than 1000 nt in length found in MNR, and the shortest length of one endornavirus found in MNR is 209 nt, which indicated that most contigs related to endornaviruses found in MNR could not be used to predict a complete ORF. This might confuse our understanding of the diversity of mycoviruses associated with MNR. To weaken this confusion, it is necessary to assess the optimal number of strains used to establish sequencing libraries and thus ensure the integrity of sequencing results.
The family Mitoviridae was newly established according to the ICTV report 2019, which contained a single +ssRNA genome with approximately 3000 nt in length, and could not assemble virus particles [70,71]. Phylogenetic analysis showed that mitoviruses clustered into three clades (clades I, II, and III), and most mitoviruses discovered in Rhizoctonia clustered into clade II, such as Rhizoctonia solani mitovirus 2, Rhizoctonia solani mitovirus 11, Rhizoctonia solani mitovirus 22, Rhizoctonia solani mitovirus 34, and Rhizoctonia solani mitovirus 35. In this study, 26 and 28 mitoviruses whose lengths are longer than 3000 nt or which can encode complete RdRp were found in BNR and MNR, respectively. Except for Rhizoctonia solani mitovirus 1-BNR (Contig2511) which was clustered into clade I, the remaining 53 putative mitoviruses were clustered into clade II. Collectively, mitoviruses found in Rhizoctonia might evolve together with their host fungi, and thus major sequence divergence in mitoviral genomes might not be induced [20,72].
Our study expands the acknowledgment of the diversity of mycoviruses present in BNR and MNR and provides the resources for investigating the evolutionary relationship of mycoviruses detected in BNR and MNR. The influence of mycoviruses found in this study on their hosts, and the interactions of mycoviruses and their host fungi need to be studied further.