Journal of General Virology (2015), 96, 714–725

DOI 10.1099/vir.0.000003

Diagnosis and discovery of fungal viruses using deep sequencing of small RNAs Eeva J. Vainio,1 Jaana Jurvansuu,1 Janne Streng,2 Minna-Liisa Rajama¨ki,2 Jarkko Hantula1 and Jari P. T. Valkonen2 Correspondence

1

Eeva J. Vainio

2

[email protected]

Received 12 September 2014 Accepted 20 October 2014

Finnish Forest Research Institute, Vantaa Research Unit, Jokiniemenkuja 1, Vantaa, Finland Department of Agricultural Sciences, P.O. Box 27, University of Helsinki, FI-00014 Helsinki, Finland

Analysis of virus-derived small RNAs with high-throughput sequencing has been successful for detecting novel viruses in plants and invertebrates. However, the applicability of this method has not been demonstrated in fungi, although fungi were among the first organisms reported to utilize RNA silencing. Here, we used virus-infected isolates of the fungal species complex Heterobasidion annosum sensu lato as a model system to test whether mycovirus genome segments can be detected with small RNA deep sequencing. Species of the genus Heterobasidion are some of the most devastating forest pathogens in boreal forests. These fungi cause wood decay and are commonly infected with species of the family Partitiviridae and the unassigned virus species Heterobasidion RNA virus 6. Small RNA deep sequencing allowed the simultaneous detection of all eight double-stranded RNA virus strains known to be present in the tested samples and one putative mitovirus species (family Narnaviridae) with a single-stranded RNA genome, designated here as Heterobasidion mitovirus 1. Prior to this study, no members of the family Narnaviridae had been described as infecting species of Heterobasidion. Quantification of viral double- and single-stranded RNA with quantitative PCR indicated that co-infecting viral species and viruses with segmented genomes can be detected with small RNA deep sequencing despite vast differences in the amount of RNA. This is the first study demonstrating the usefulness of this method for detecting fungal viruses. Moreover, the results suggest that viral genomes are processed into small RNAs by different species of Heterobasidion.

INTRODUCTION RNA interference (RNAi), also called RNA silencing, refers to post-transcriptional control of gene expression and protection of organisms against mobile repetitive elements, retrotransposons and viral infections (Tijsterman et al., 2002). The cellular machinery for RNAi is present in animals, plants and fungi (Meister & Tuschl 2004; Chang et al., 2012). During RNA silencing, RNase III endoribonucleases called Dicers (or Dicer-like in plants) detect and cleave double-stranded RNA (dsRNA) templates into small interfering RNAs (siRNAs), which are typically 21, 22 or 24 nt long in plants (Bologna & Voinnet, 2014). The siRNAs then associate with an RNA-induced silencing complex in which a protein of the Argonaute family catalyses the sequence-guided degradation of target mRNA. Because the RNAi machinery targets potentially deleterious non-self nucleic acids, virus-infected host organisms are typically enriched for siRNAs for viral sequences. The GenBank/EMBL/DDBJ accession number for the complete genome sequence of HetMV1-an1 is KJ873059. Three tables are available with the online Supplementary Material.

714

Consequently, the analysis of virus-derived small RNAs with high-throughput sequencing has been successfully used in plants and invertebrates to detect novel viruses (Kreuze et al., 2009; Pantaleo et al., 2010; Wu et al., 2010; Bi et al., 2012; Kashif et al., 2012). Among fungi, RNAi was first observed in the ascomycetous Neurospora crassa (Romano & Macino, 1992); the transformation of N. crassa with multiple adjacent copies of genes required for carotenoid pigment biosynthesis unexpectedly results in reduced pigmentation, a phenomenon the authors called ‘quelling’. Segers et al. (2007) demonstrated for the first time that RNAi acts as an antiviral defence mechanism in fungi by showing that Dicer gene disruptions in the ascomycetous Chestnut blight fungus, Cryphonectria parasitica, result in increased viral RNA levels. The loss of RNAi in the budding yeast, Saccharomyces cerevisiae (phylum Ascomycota), is believed to result from adaptation to the presence of Saccharomyces cerevisiae L-A virus. This virus and its satellite-encoded ‘killer’ toxins provide a selective advantage for the host in competition against yeast strains lacking immunity to the toxin (Drinnenberg et al., 2009, 2011).

Downloaded from www.microbiologyresearch.org by 000003 G 2015 The Authors IP: 37.44.207.73 On: Sat, 28 Jan 2017 02:49:43

Printed in Great Britain

Deep sequencing of mycoviral small RNAs

The Heterobasidion annosum sensu lato species complex includes some of the most devastating forest pathogens in boreal forests. These fungi spread efficiently via basidiospores that infect fresh stump surfaces and butt wounds of living trees (Stenlid & Redfern, 1998). They also spread vegetatively via root contacts and single clonal individuals can survive for decades and infect dozens of trees. According to their host tree preferences, and genetic evidence, species of H. annosum s. lat. can be divided into two major clades: the ‘spruce clade’ includes the Eurasian species Heterobasidion parviporum and the Southern European Heterobasidion abietinum together with the North American Heterobasidion occidentale, whereas the European Heterobasidion annosum and the North American Heterobasidion irregulare form the ‘pine clade’. The recently published genome of H. irregulare contains the core components of the RNAi machinery, such as three Dicer-encoding genes and seven Argonaute-encoding genes (Olson et al., 2012; Hu et al., 2013). dsRNA virus infections occur in approximately 15–17 % of cultured Heterobasidion strains (Ihrmark, 2001; Vainio et al., 2011b). Most of these infections do not cause drastic phenotypic effects on their hosts when cultured in the laboratory, but some affect the growth of the host fungus (Vainio et al., 2010, 2012) or may reduce the germination frequency of basidiospores (Ihrmark et al., 2004). Recently, we showed that a single Heterobasidion-infecting virus strain may have beneficial, commensal or detrimental effects on a single host isolate depending on culture conditions and interacting antagonistic fungi (Hyder et al., 2013). Viruses that infect Heterobasidion species are transmitted between host strains during anastomosis of fungal hyphae, which also allows virus transmission between incompatible fungal strains and distantly related species of Heterobasidion (Ihrmark, 2001; Vainio et al., 2010, 2011a, b). Known viruses of Heterobasidion also spread via basidiospores and conidia (Ihrmark et al., 2002, 2004). Heterobasidion RNA virus 6 (HetRV6) is responsible for ~70 % of the dsRNA virus infections among Heterobasidion strains in Europe (Vainio et al., 2012). The RNAdependent RNA polymerase (RdRp) gene of HetRV6 shares a moderate level of sequence similarity with the Curvularia thermal tolerance virus (CThTV), which is involved in a three-way symbiosis with its host fungus (the endophytic ascomycete Curvularia protuberata) and a grass species (Ma´rquez et al., 2007). Whether HetRV6 encompasses a second genomic segment similar to CThTV is unknown. Members of the virus family Partitiviridae infect all five species of H. annosum s. lat. (Ihrmark, 2001; Vainio et al., 2010, 2011a, b, 2013b). Partitiviruses have bisegmented dsRNA genomes that encode an RdRp and a capsid protein (CP). The two genomic RNAs are separately encapsidated (King et al., 2012; Nibert et al., 2013). The RdRp of partitiviruses mediates plus-strand RNA synthesis (transcription) inside the virus particle using the minus strand of the genomic dsRNA as a template, after which the genomic plus strand is extruded for use in translation in http://vir.sgmjournals.org

the host cytosol (Nibert et al., 2014). Mitochondrial viruses (genus Mitovirus) of the family Narnaviridae are common in fungi (Hillman & Cai, 2013), but only two mitovirus species have been described in basidiomycete fungi and no mitoviruses are known to infect Heterobasidion spp. Fungal mitoviruses are unencapsidated viruses with positive-sense single-stranded RNA (ssRNA) genomes usually predicted to contain only a single gene that encodes RdRp (King et al., 2012). Cryphonectria parasitica strains with Dicer gene disruptions have been used to demonstrate that RNAi targets the hypoviruses and mycoreoviruses of Cryphonectria parasitica (Segers et al., 2007; Zhang et al., 2008, 2012) as well as a partitivirus and a victorivirus that naturally infect the ascomycetous root rot pathogen Rosellinia necatrix (Chiba et al., 2013a, b). Hypoviruses have positive-sense ssRNA genomes and cause hypovirulence in Cryphonectria parasitica, whereas mycoreoviruses, partitiviruses and victoriviruses have dsRNA genomes (King et al., 2012). To our knowledge, RNAi against the HetRV6/CThTV-like virus clade or mitoviruses has not been demonstrated. Our aim was to test the following hypotheses: (i) small RNA deep sequencing can be used for virus detection in fungi, (ii) Heterobasidion species process viral RNAs into siRNAs and (iii) small RNA deep sequencing can be used to discover previously unknown viral species from Heterobasidion spp.

RESULTS Yield and size distribution of the siRNA reads The small RNA deep sequencing analysis included small RNA from ten fungal strains representing four different species of Heterobasidion. Seven of these strains were known to be infected with species of the family Partitiviridae or HetRV6 (Table 1). The total number of reads produced by the Illumina sequencer was 27 014 983 (read length 10– 34 nt), corresponding to 1351 megabases. Most clusters (92 %) fulfilled the default Illumina quality criteria and 96 % of bases had a quality score ¢30. A relatively high proportion (49 %) of the reads consisted of RNAs of 10– 17 nt, whereas reads of 18–26 nt accounted for 43 % of the RNAs. The two most prominent read lengths were 13 nt and 19 nt (Fig. 1). Sequence contigs produced with Velvet The Velvet program was used for de novo assembly of contigs using the small RNA reads. Contig assembly was first conducted using reads of 21–24 nt, which are typical lengths for siRNAs generated by many Dicer proteins (Chang et al., 2012; Bologna & Voinnet, 2014). However, because a considerable proportion of reads were shorter than 21 nt (Fig. 1), we also examined the presence of viral sequences among the shorter siRNAs. Viral contigs were obtained with Velvet using reads of 18 nt and larger, and therefore we included all reads of 18–27 nt in the

Downloaded from www.microbiologyresearch.org by IP: 37.44.207.73 On: Sat, 28 Jan 2017 02:49:43

715

E. J. Vainio and others

Table 1. Heterobasidion isolates used for small RNA deep sequencing Species

Strain

Virus infection(s)*

H. abietinum H. abietinum H. annosum

04188 RF2s2 94233

HetRV6-ab6 Not known HetPV13-an1

H. annosum H. irregulare H. parviporum

05021 704 7R18

HetMV1-an1 Not known HetPV17-ir1 HetPV2-pa1

H. parviporum

7R226

HetPV7-pa1

H. parviporum H. parviporum

RT3.49C 95122

HetRV6-pa19-c HetPV4-pa1 HetPV15-pa1

H. parviporum

10084

Not known

GenBank

Taxonomy

ReferenceD

HQ189459 – KF963177 (RdRp) KF963178 (CP) KJ873059 – KJ873060 HM565953 (RdRp) HM565954 (CP) JN606091 (RdRp) JN606090 (CP) KF551892 HQ541325 (RdRp) KF963186 (RdRp) KF963187 (CP) –

Unassigned – Alphapartitivirus

Vainio et al., (2012) – Kashif et al., (2015)

Mitovirus – Alphapartitivirus Betapartitivirus

This study – E. J. Vainio and others, unpublished Vainio et al., (2011a)

Betapartitivirus

Vainio et al., (2014)

Unassigned Alphapartitivirus Alphapartitivirus

Vainio et al., (2011b, 2013a) Kashif et al., (2014)

–

–

*Note that the nomenclature of partitiviruses has recently changed. HetPV2 and HetPV4 were formerly designated HetRV2 and HetRV4, respectively (Nibert et al., 2014). DAll other virus strains except HetMV1-an1 were known to be present in the tested fungal strains, whereas HetMV1-an1 was newly identified with small RNA deep sequencing in this study.

subsequent Velvet analyses. Using k-mer values smaller than 13 or larger than 19 resulted in significant decreases in contig number (the k-mer value indicates the word size used in the de Bruijn graphs) (Zerbino & Birney, 2008). We selected a k-mer value of 13 to be used in an optimized Velvet run in which host rRNA sequences were subtracted from the sequence pool using the Coderestar script and the primary contigs were then post-assembled with the AssemblyAssembler program into larger contigs. This optimized Velvet run resulted in 1608 contigs 56–515 nt in

Percentage of all reads

15

10

5

0 10

20 Read length (nt)

30

Fig. 1. Size distribution of small RNA reads obtained using Illumina deep sequencing. 716

length and showed that viral contigs accounted for 3.7 % (n560) of all contigs detected. Using k-mer values of 15, 17 and 19, the number of contigs produced was 1421, 747 and 114, and the largest contigs were 544, 765 and 584 nt in size, respectively. All Velvet assemblies that produced any contigs were examined for the presence of viral sequences using BLASTN and BLASTX. Identification of viral sequences with

BLAST

The contigs produced with Velvet were used to search similar sequences in the NCBI/GenBank database at the nucleotide and protein level using BLASTN and BLASTX. This analysis revealed HetPV2-pa1 and HetPV7-pa1 sequences encoding RdRp and CP; and HetPV4-pa1, HetPV13an1 and HetRV6-pa19-c sequences encoding RdRp. Heterobasidion partitivirus 2 (HetPV2) and HetPV7 were recently classified by the International Committee on the Taxonomy of Viruses into the genus Betapartitivirus (Nibert et al., 2014) and share 65.5 % polymerase sequence identity at the nucleotide level. The remaining partitiviruses included in the analysis represent the genus Alphapartitivirus and share 39.7–65.6 % polymerase sequence identity at the nucleotide level (Table S1, available in the online Supplementary Material). In addition to these viruses that are known to infect the fungal strains used in our study, we found some contigs that resembled fungal mitoviruses. In the optimized Velvet run, we found nine mitoviral contigs 58–241 nt in length that had 42–94 % sequence identity with known mitoviruses, as determined using BLASTX.

Downloaded from www.microbiologyresearch.org by IP: 37.44.207.73 On: Sat, 28 Jan 2017 02:49:43

Journal of General Virology 96

Deep sequencing of mycoviral small RNAs

Mitovirus sequence characterization and phylogenetic analysis

Guide strand-mediated assembly with the MAQ sequence mapping and assembly program

The short (58–241 nt) Velvet contigs with sequence similarity to mitoviruses were assembled using previously described mitoviral genome sequences at the nucleotide and protein levels to reveal their putative genomic locations and to design PCR primers for virus screening. The cDNAs of the ten Heterobasidion strains used in this study were screened one at a time using mitovirus primers and revealed the presence of a new putative mitovirus, designated here as HetMV1-an1 (Heterobasidion mitovirus 1, strain 1 from H. annosum), in H. annosum 94233. Subsequently, the intervening sequence regions that were not elucidated with small RNA analysis were amplified with PCR using specific primers and the products were sequenced. The resulting complete genome sequence of HetMV1-an1 (4379 nt in length) was deposited in GenBank (accession number KJ873059). Mitoviral amplification products were not detected using total cellular nucleic acids (including nuclear and mitochondrial DNA) as a PCR template and therefore HetMV1-an1 appeared not to be incorporated into the host genome, in contrast to a mitovirus described in the basidiomycete fungus Thanatephorus cucumeris (anamorph Rhizoctonia solani; Table 2) (Lakshman et al., 1998).

The MAQ program was used to map small RNA reads to reference sequences, namely all characterized viral genome segments found previously in the Heterobasidion strains used in our study (Table 1). Small RNAs covered 90.2 % and 99.3 % of the polymerase-encoding genome segments and 97.3 % and 96.6 % of the capsid-encoding segments of the two betapartitiviruses, HetPV2-pa1 and HetPV7-pa1, respectively. The average sequencing depth for these segments varied from 466 to 3106 (Table 3). The polymerase-encoding segments of HetRV6-ab6, HetRV6pa19-c, HetPV4-pa1, HetPV13-an1 and HetMV1-an1 were also well represented among small RNAs (90.2–98.0 % sequence coverage and sequencing depth of 14–1476; Table 3). However, virus strains HetPV15-pa1 and HetPV17-ir1 were less well covered by small RNAs (sequence coverage 28.8–38.1 %) and their MAQ scores were less reliable (Table 3). Moreover, sequencing depth was very low for these two viruses (Table 3) and they were not detected by Velvet. In all cases, the distal ends of the genome segments seemed to be poorly detected by MAQ. Nevertheless, all viruses known to be present in the sample were detected by MAQ.

The host of HetMV1-an1 (H. annosum 94233) is co-infected with the partitivirus HetPV13-an1 (GenBank accession numbers KF963177 and KF963178). However, we did not detect mitoviral dsRNA during genome characterization of the co-infecting partitivirus (Kashif et al., 2014). Mitoviruses are positive-sense ssRNA viruses, but their replicative intermediates occur in double-stranded form and they can often be detected as dsRNA in the presence of a sufficiently high copy number (Tuomivirta & Hantula 2003; Martı´nezA´lvarez et al., 2014). Sequence comparisons (Table 2) and phylogenetic analysis (Fig. 2) indicated that HetMV1-an1 is a novel putative mitovirus species that shares approximately 33 % identity at the protein level with its closest relative, Clitocybe odora mitovirus (Heinze, 2012). BLASTP comparisons revealed sequence identities of 32–48 % with other mitoviral species (Table 2), but this analysis recognized only sequence regions that overlapped with the conserved mitovirus RdRp protein family domain (pfam05919; aa residues 339–652 in the predicted RdRp of HetMV1-an1). At the nucleotide level, BLASTN recognized only short nucleotide stretches of related mitoviruses, most covering less than 1 % of the HetMV1-an1 genome (Table 2). The 4379 nt HetMV1-an1 sequence has a DNA G+C content of 40.0 mol%. The genome of HetMV1-an1 contains an ORF encoding RdRp of 836 aa when translated using mitochondrial codon usage (translation table 4). This ORF is interrupted by nine UGA stop codons when translated using the standard nuclear translation table (Fig. 3). No other long ORFs were detected. The lengths of the 59- and 39-untranslated regions were 801 nt and 1067 nt, respectively. http://vir.sgmjournals.org

The CP-encoding segment of HetPV13-an1 seemed to be less well covered by small RNAs than the RdRp gene (sequence coverage 76.4 % and 96.0 %, respectively). Moreover, the MAQ scores were less reliable for the CP segment compared with RdRp (Table 3), suggesting that the two genome segments may have differed with respect to sheer quantity. This idea was supported by quantitative PCR (qPCR) analysis, which showed that the RdRp:CP ratio in HetPV13-an1 was 12.0 for dsRNA and 8.7 for viral RNA (consisting mainly of ssRNA; Fig. 4). The opposite situation was observed for HetPV2-pa1, which seemed to produce more CP-derived small RNA than RdRpderived small RNA (sequence coverage 97.3 % and 90.2 %, respectively). The presence of two highly similar HetRV6 strains that share 92.5 % sequence identity (HetRV6-ab6 and HetRV6-pa19-c) resulted in some ambiguity in the MAQ base calls (for example, sequence positions differing for C or T between the two sequences resulted in a Y base call), presumably owing to cross-assembly. On the other hand, clear signature nucleotides were observed that indicated the presence of both virus strains among the siRNAs (Table S2). In the case of the partitiviruses, polymerase-encoding nucleotide sequence identity among strains included in the sample pool was moderate or low (39.7–65.6 %; Table S1). The CP-encoding sequences of HetPV4-pa1 and HetPV17ir1 have not been determined and whether HetRV6 has a CP is not known; we therefore attempted to identify these genome segments using CP gene sequences from the most closely related viruses (as determined with RdRp sequence comparisons using NCBI BLASTN) for MAQ runs (Table S3). However, the analysis yielded only low-quality sequence matches and, despite testing several PCR primers,

Downloaded from www.microbiologyresearch.org by IP: 37.44.207.73 On: Sat, 28 Jan 2017 02:49:43

717

E. J. Vainio and others

Table 2. Sequence similarity of HetMV1-an1 compared with related mitoviruses as determined with NCBI alignments Virus species (RdRp) Clitocybe odora virus Ophiostoma mitovirus 3b Botrytis cinerea mitovirus 1 Ophiostoma mitovirus 1b Tuber aestivum mitovirus Ophiostoma mitovirus 3a Sclerotinia homeocarpa mitovirus Sclerotinia sclerotiorum mitovirus 3 Ophiostoma mitovirus 1a Sclerotinia sclerotiorum mitovirus 6 Sclerotinia sclerotiorum mitovirus 14 Thanatephorus cucumeris mitovirus

BLASTP identity/ query cover (%)*

BLAST

and sequence

Accessions (nt, aa)

Length (nt)

identity/ Overall identity query cover (%) according to MAFFTD

HE717021.1, YP_005352912.1 AM087550, CAJ32468.1 EF580100, ABQ65153.3 AM087549, CAJ32467.1 HQ992989.1, YP_004564622.1 AJ004930.1, NP_660176.1 AY172454, AAO21337.1

3765 2332 2804 2572 3480 2617 2632

53/62 45/42 45/41 44/45 44/50 48/42 47/46

68/13

NS

32.7 % 27.4 % 26.7 % 25.2 % 26.8 % 25.9 % 25.9 %

JX401537, AGC24232.1

2588

46/28

NS

23.6 %

AM087548, CAJ32466.1 KJ462513, AHX84133.1

3147 2586

36/51 40/44

NS NS

24.4 % 24.9 %

KF913889, AHF48630.1

2564

40/44

NS

24.8 %

TCU51331, AAD17381.1

3570

32/58

NS

24.1 %

BLASTN

NS NS NS NS NS

NS, No significant similarity (query cover less than 5 %). *Query cover was identified with BLASTP and BLASTN and indicates the proportion (%) of the complete HetMV1-an1 sequence of 4379 nt (KJ873059) that was aligned with each particular reference sequence available in GenBank. The e-values for BLASTP comparisons were 7276–5244 and the BLASTN e-value for Clitocybe odora virus was 8215. DOverall sequence identity in a sequence alignment produced by MAFFT using the complete sequence for HetMV1-an1.

we obtained no amplification products of expected sizes (data not shown). This might have been expected as the CP-encoding segments of partitivirus species typically share lower sequence identity compared with the corresponding RdRp-encoding segments (Nibert et al., 2014), which shared 68–73 % sequence identity in our analysis (Table S3). Alternatively, the CP-encoding segments of HetPV4-pa1 and HetPV17-ir1may not be subjected to RNA silencing.

H. annosum 94233 is co-infected with HetMV1-an1 and the partitivirus strain HetPV13-an1, which showed small RNA sequence coverages of 98 % and 96 %, respectively, for the polymerase gene. Based on qPCR, 2.3 times more mitoviral dsRNA was present compared with partitiviral dsRNA in the infected fungal strain. However, on average, 200 and 1780 times more HetMV1-an1 ssRNA than HetPV13-an1 RdRp and CP transcripts, respectively, were present (Fig. 4).

Amounts of viral RNAs in co-infected isolates

DISCUSSION

Two Heterobasidion isolates with viral co-infections were used to examine the relative amounts of viral dsRNA and ssRNA present in the host mycelia. The double-infected H. parviporum 7R226 isolate harbours the partitivirus HetPV7pa1 and the unassigned HetRV6-pa19-c. The polymerase genes for HetRV6-pa19-c and HetPV7-pa1 were both well represented among the small RNAs (sequence coverage 94.6 % and 99.3 %, respectively). However, based on qPCR, 5.9 times more HetPV7-pa1 polymerase dsRNA was present compared with HetRV6-pa19-c polymerase dsRNA in the co-infected host isolate (Fig. 5). Notably, substantial differences were apparent in viral transcription activity, as 19.7 times more HetPV7-pa1 RdRp transcripts (i.e. viral RNA including mainly ssRNA) were present than HetRV6pa19-c RdRp transcripts (Fig. 5). 718

This study demonstrates that small RNA deep sequencing can be used to detect fungal viruses and suggests that Heterobasidion species process viral RNA into small RNAs. The method allowed the simultaneous detection of nine mycoviral strains with dsRNA or ssRNA genomes, including a new putative mitovirus. The sequence of this previously unknown mitovirus shared only ~33 % sequence similarity with its closest relative, the Clitocybe odora mitovirus (Heinze, 2012), indicating that relatively divergent viral species can be detected with small RNA deep sequencing. However, as the identification of viral contigs is based on the availability of reference sequences in public databases, this method cannot be used to detect viruses with no known relatives. Hence, small RNA-based diagnostics of viruses benefit from the accumulation of sequence data of novel

Downloaded from www.microbiologyresearch.org by IP: 37.44.207.73 On: Sat, 28 Jan 2017 02:49:43

Journal of General Virology 96

Deep sequencing of mycoviral small RNAs

0.99 1 1 1

1 1 1

Saccharomyces 23S RNA narnavirus (AAC98708) Cryphonectria parasitica mitovirus 1-NB631 (NP_660174.1) 1 HetMV1-an1 (AIF33766) Clitocybe odora virus (YP_005352912.1) Tuber aestivum mitovirus (YP_004564622.1) 1 Ophiostoma mitovirus 1b (CAJ32467.1) 1 Ophiostoma mitovirus 1a (CAJ32466.1) Thanatephorus cucumeris mitovirus (AAD17381.1) 1 Botrytis cinerea mitovirus 1 (ABQ65153.3) Ophiostoma mitovirus 3b (CAJ32468.1) Sclerotinia sclerotiorum mitovirus 3 (AGC24232.1) 1 1 Sclerotinia sclerotiorum mitovirus 14 (AHF48630.1) 1 Sclerotinia sclerotiorum mitovirus 6 (AHX84133.1) 1 Sclerotinia sclerotiorum mitovirus 12 (AHF48628.1) 1 Ophiostoma mitovirus 3a (NP_660176.1) Sclerotinia homeocarpa mitovirus (AAO21337.1) Fusarium circinatum mitovirus 2-1 (AHI43534) Gremmeniella abietina mitovirus 1 (YP_077184.1) Fusarium circinatum mitovirus 1 (AHI43533) 1 1 Thielaviopsis basicola mitovirus (YP_002822229.1) 1 Ophiostoma mitovirus 4 (NP_660179.1) 1 Sclerotinia sclerotiorum mitovirus 4 (AGC24233.1) 1 1 Ophiostoma mitovirus 5 (NP_660180.1) Sclerotinia sclerotiorum mitovirus 1 (AEX91878.1) Gremmeniella abietina mitovirus 2 (AEY76153.1) 1 Ophiostoma mitovirus 6 (NP_660181.1) Helicobasidium mompa mitovirus (BAD72871.1) Cryphonectria cubensis mitovirus 1a (AAR01970.1) Sclerotinia sclerotiorum mitovirus 2 (AEX91879.1) Tuber excavatum mitovirus (AEP83726.1) 0.3

Fig. 2. Bayesian analysis of mitovirus polymerase sequences including HetMV1-an1 and a selection of associated taxa. Bar, 0.3 amino acid substitutions per site; numbers at branch nodes show percentage posterior probabilities using the substitution model Blosum (only values ¢0.95 are shown).

mycoviral species in public databases (Osaki et al., 2006; Chiba et al., 2009; Feldman et al., 2012; Lin et al., 2012; Yaegashi et al., 2013). Recent next-generation sequencing studies have used purified dsRNA as a template and detected mycoviral sequences in grapevine plants (Coetzee et al., 2010; Al Rwahnih et al., 2011) and endophytic fungi (Feldman et al., 2012). Direct mRNA sequencing has revealed novel virus species in plants (Al Rwahnih et al., 2009; Wylie & Jones, 2011; Wylie et al., 2013) and fungi (Schoebel et al., 2014). Compared with mRNA sequencing, small RNA deep sequencing enriches the sequence pool for viral sequences and the advantage of this method over high-throughput sequencing of dsRNA is that

~4.4 kb

5′

3′

Fig. 3. Genome organization of HetMV1-an1. The 59- and 39untranslated regions are shown in black. Arrows indicate the positions of UGA codons within the predicted RNA-dependent RNA polymerase (RdRp)-encoding region. UGA is a stop codon when translated by standard codon usage, but encodes tryptophan using the mitochondrial translation table. The core RdRp domain (aa residues 339–652 of the 836 aa ORF) is shown in grey. http://vir.sgmjournals.org

small RNA deep sequencing may allow detection of a wider variety of viruses, including those with ssRNA or DNA genomes (Kreuze et al., 2009; Kashif et al., 2012). The novel putative mitovirus detected in this study has an ssRNA genome and remained undetected using a conventional cloning approach (Kashif et al., 2014). Among fungi, only one virus species with a DNA genome has been described (Yu et al., 2010), and whether the RNA transcripts of fungal DNA viruses are processed into small RNAs by the hosts remains unknown. In this study, the sequence-guided assembly by MAQ revealed high small RNA coverage for most viruses known to be present in the tested fungal strains. The genome of the betapartitivirus HetPV7-pa1 was almost fully covered (sequence coverages of 99.3 % and 96.6 % for the RdRpand CP-encoding segments, respectively). The genomes of HetRV6-ab6, HetRV6-pa19-c, HetPV4-pa1 and HetPV2pa1 were also well represented among small RNAs (90.2– 97.3 % sequence coverage). However, HetPV15-pa1 and HetPV17-ir1 showed sequence coverages of only 28.8– 38.1 %. In addition to viral copy number, differences in the amount of small RNAs from different virus species may be related to the regulation of RNAi by the host or virus. Recent studies have shown that suppression of RNAi is a common phenomenon among fungal viruses. Hammond

Downloaded from www.microbiologyresearch.org by IP: 37.44.207.73 On: Sat, 28 Jan 2017 02:49:43

719

E. J. Vainio and others

720

Table 3. Coverage of small RNAs determined with the MAQ sequence mapping and assembly program Virus strain

RNA-dependent RNA polymerase (RdRp) Reference sequence length (nt)

HetPV2-pa1 HetPV4-pa1 HetPV7-pa1 HetPV13-an1 HetPV15-pa1 HetPV17-ir1 HetRV6-ab6 HetRV6-pa19-c HetMV1-an1

2290 2007 2297 1873 1882 2016 2050 1957 4379

Capsid protein (CP)

Coverage (%)*

High quality coverage (%)D

Depth (¾)d

Reference sequence length (nt)

Coverage (%)*

High quality coverage Depth (¾)d (%)D

90.2 93.6 99.3 96.0 28.8 29.8 91.6 94.6 98.2

48.6 39.4 88.9 19.5 0.9 1.0 41.0 48.9 67.4

45.5 23.2 310.4 14.3 1.0 6.5 24.5 28.3 146.6

2238

97.3

63.9

NA

NA

NA

NA

2231 1776 1791

96.6 76.4 38.1

72.5 1.7 0.9

78.1 3.3 1.2

49.0

NA

NA

NA

NA

ND

ND

ND

ND

ND

ND

ND

ND

2

2

2

2

NA,

Not available – the CP sequence has not been described. No data – a putative CP has not been detected for strains of HetRV6; 2, The mitovirus genome does not include a CP. *Sequence coverage is the percentage sequence overlap determined by aligning the MAQ output sequences, indicating sites covered by siRNAs, with the original reference sequence. DThe proportion (%) of the mapped sequence with a quality score of ¢90 %. dAverage sequencing depth across all non-gap regions as determined by MAQ. ND,

Journal of General Virology 96 Downloaded from www.microbiologyresearch.org by IP: 37.44.207.73 On: Sat, 28 Jan 2017 02:49:43

Deep sequencing of mycoviral small RNAs

(a)

3 2.5 2 1.5 1 0.5 HetMV1

(b)

600 500 400 300 200 100 5

0

8

6

4

2

0

HetPV13-an1 RdRp HetPV13-an1 CP

Viral RNA/GAPDH mRNA

Viral RNA/GAPDH mRNA

Viral dsRNA/HetRV6-pa19-c RdRP dsRNA

Viral dsRNA/HetPV13-an1 RdRp dsRNA

3.5

0 (b)

(a)

4

HetPV13-an1 RdRp HetPV13-an1 CP

Fig. 4. The amount of viral dsRNA (a) and viral RNA (mainly ssRNA) (b) in H. annosum 94233. Viral dsRNA and total RNA were isolated from separate fungal cultures. To accommodate for possible variation in virus copy number among fungal mycelia, the number of replicate cultures was three for each experiment. Error bars show standard error. Note the interrupted scale for viral RNA.

et al. (2008) showed that RNAi in Aspergillus nidulans is suppressed by Aspergillus virus 1816 and Aspergillus virus 341 is targeted for degradation by RNAi by the same host. Similarly, Chiba et al. (2013b) showed that a victorivirus (family Totiviridae) of R. necatrix is targeted by RNAi and a mycoreovirus infecting the same host suppresses RNAi (Yaegashi et al., 2013). The siRNA read length profile observed in this study differed from a characteristic profile produced by Dicerlike RNases and a significant proportion of the small RNAs were shorter than the typical small RNAs of 21–25 nt produced by the ascomycetes N. crassa and Cryphonectria parasitica or the zygomycete Mucor circinelloides (Chang et al., 2012). The functionality of the three putative Dicerencoding genes and seven Argonaute-encoding genes identified in the genome of H. irregulare has not been demonstrated experimentally. In Cryphonectria parasitica, the most prominent size class of hypovirus-derived siRNAs is 20–22 nt, but viral siRNAs are also present among siRNAs of 19 nt and 23–28 nt (Zhang et al., 2008). In this study, two of the Heterobasidion host strains were co-infected with distantly related virus species. Quantification of viral dsRNA and ssRNA indicated that http://vir.sgmjournals.org

HetPV7-pa1 RdRp

HetRV6-pa19-c RdRp

HetPV7-pa1 RdRp

3

2

1

0 HetMV1-an1

HetRV6-pa19-c RdRp

Fig. 5. The amount of viral polymerase dsRNA (genome segments) (a) and viral RNA (mainly ssRNA) (b) in H. parviporum 7R226. Viral dsRNA and total RNA were isolated from separate fungal cultures. To accommodate for possible variation in virus copy number among fungal mycelia, the number of replicate cultures was three for each experiment. Error bars show standard deviation.

co-infecting viral species can be detected with small RNA deep sequencing despite vast differences in the amount of RNA. In the case of HetMV1/HetPV13 co-infection, we found over 200 times more mitovirus RNA than partitivirus RdRp RNA, although the ratio of HetMV1 : HetPV13 dsRNA was only 2.3. The mitoviral coding strand is present in infected tissue in a greater molar amount than the dsRNA, which is considered to be the replicative form of these positive-sense ssRNA viruses (King et al., 2012). The HetMV1-an1 RdRp-encoding sequence contains nine internal stop codons when translated using the standard nuclear translation table. Therefore, the mitoviral coding strand is predicted to be translatable only in mitochondria by mitochondrial codon usage. In the case of HetRV6/ HetPV7 co-infection, we also found a significant difference in the amount of viral polymerase transcripts (ssRNA) and genomic RNA from the two co-infecting viruses. Whether co-infecting Heterobasidion viruses have additive, synergistic or antagonistic interactions in the co-infected hosts remains unclear. Co-infection may affect genome copy numbers, as has been demonstrated among many plant pathogenic viruses (Poolpol & Inouye,1986; Mukasa et al., 2006).

Downloaded from www.microbiologyresearch.org by IP: 37.44.207.73 On: Sat, 28 Jan 2017 02:49:43

721

E. J. Vainio and others

We observed differences in the siRNA accumulation between the two genome segments of the partitiviruses, suggesting that the quantity of the two RNAs differs or that they are differentially targeted by RNAi. In particular, the CP-encoding RNA segment of HetPV13-an1 appeared to be less well represented in the small RNA pool compared with the RdRp segment. This quantitative discrepancy between HetPV13-an1 genome segments was supported by qPCR analysis. Moreover, Jurvansuu et al. (2014) found that the ratio of RdRp- and CP-encoding dsRNA for HetPV2-pa1 is on average 0.55, which is in accordance with the results of our study, suggesting that CP of HetPV2-pa1 is better represented in the siRNA pool compared with RdRp. Taken together, this study shows that Heterobasidion species are able to process the RNA of various viruses into small RNAs; in addition to partitiviruses these viruses include the unassigned taxon HetRV6 and the new putative species of the genus Mitovirus. This observation is noteworthy as the entire reproductive cycle of mitoviruses is thought to occur inside host mitochondria (King et al., 2012; Hillman & Cai, 2013). Thanatephorus cucumeris mitovirus, which moderately resembles HetMV1-an1 (~24 % sequence identity at the protein level), is integrated into the host genome (Lakshman et al., 1998), but this appears not to be the case for HetMV1-an1. No members of the virus family Narnaviridae have been described from species of Heterobasidion despite extensive screening (Vainio et al., 2010, 2011a, b, 2012, 2013a, b, 2014; Kashif et al., 2014). Thus, the power of the small RNA deep sequencing approach in screening for new viruses is apparent.

METHODS Fungal strains and viral infections. Isolates of ten strains of

Heterobasidion were used for the small RNA analysis: H. abietinum 04188 and RF2s2; H. annosum 94233 and 05021; H. irregulare 704; and H. parviporum 7R18, 10084, RT3.49C, 95122 and 7R226. Seven of these isolates were known to be infected with partitiviruses or HetRV6 (Table 1), whereas three (05021, 10084 and RF2s2) contained no known virus infections. H. parviporum 7R226 was known to be coinfected with strains of two viruses: Heterobasidion partitivirus 7 (strain HetPV7-pa1) and HetRV6 (strain HetRV6-pa19-c). Isolation and high-throughput sequencing of small RNAs.

Fungal mycelia were cultivated on modified orange serum agar plates covered with cellophane membranes. Hyphae were collected, frozen with liquid nitrogen and ground with a mortar and pestle. Total RNA was isolated from ~30–50 mg ground fungal mycelia using TRI Reagent (Molecular Research Center) and dissolved in sterile water that had been treated with diethylpyrocarbonate. The quantity and quality of the RNA samples were measured with agarose gel electrophoresis and a spectrophotometer (Qubit fluorometer; Invitrogen) and equal amounts of total RNA from ten Heterobasidion isolates were pooled. The quality of the pooled RNA sample (5 mg) was verified by the Fasteris SA Company, Switzerland, using a BioAnalyser 2100 (Agilent Technologies). The pooled sample had an rRNA (28S : 18S) ratio of ¢2.6 and RNA integrity number of ¢6.9. The pooled RNA sample was used for siRNA library construction by Fasteris SA, followed by high-throughput sequencing using an Illumina HiSeq 2500. Briefly, the sample processing included 722

acrylamide gel purification of small RNA fragments, single-stranded ligation of 39- and 59-adapters, reverse transcription and PCR amplification to generate a cDNA library. Isolation of dsRNA with cellulose affinity chromatography. The

presence of mitoviral dsRNA in H. annosum 94233 was examined with CF11 cellulose affinity chromatography (Morris & Dodds 1979; Tuomivirta et al., 2002; Jurvansuu et al., 2014) using 2.5–3.0 g (fresh mass) fungal mycelia. The protocol included isolation of total nucleic acids using phenol/chloroform extraction followed by specific precipitation of the dsRNA fraction using CF11 cellulose powder (Whatman) in 15 % (v/v) ethanol. Mitovirus sequence characterization with Sanger sequencing.

The complete genome of the novel putative mitovirus strain, HetMV1-an1, was determined with Sanger sequencing. cDNA was generated as described (Vainio et al., 2013a). Briefly, the protocol included isolation of total nucleic acids with phenol/chloroform extraction and precipitation with polyethylene glycol, followed by cDNA synthesis using random priming and the RevertAid reverse transcriptase (Thermo Scientific). PCR primers were designed based on siRNA contigs with mitoviral affinities as identified using Velvet (Zerbino & Birney, 2008; see below) and BLASTX (National Center for Biotechnology Information; www.ncbi.nlm.nih.gov). The following primers were used for sequence determination: HetMV1MidF (59-CCTATGGATGGGACTTTTGACC-39), HetMV1MidRev (59-TTTGCGAATTCAAGCCCGAT-39), MV1MidFB (59-GCCGTGCAACAAGCTATT-39), MV1MidRevC (59-CAACGTCGTCCATGCTTT-39), MV1EndF (59-GGTTGAGGTTGGTTTGGCTA-39), MV1BegEV (59-CCTATCTCGGCCAAAGACGA-39), MV1MidFC (59-ACCTGTGAGGCAACTCTGCT-39), MV1BegRev (59-ATGCCCTGCAATAGCTTGTT-39), MV1ERev2 (59-TTGAATTTCACCAGGGAAGA-39), MV1BR3 (59-AACCCTGAAAAGCGAAATCA-39), MV1EF2 (59-TCAAGGATCAAGTAAGTGGTCA-39) and MV1endRevS (59-TATGGCAGGTGAAACATGGA-39). PCR was performed using the Dynazyme DNA polymerase (Thermo Scientific) and sequencing was conducted at Macrogen, South Korea. All sequence positions were assessed in triplicate using three separate PCR analyses from two independently synthesized cDNAs. Sequence ends were determined using the single primer amplification technique of Lambden et al. (1992) as modified by Vainio et al. (2011a). The protocol includes the ligation of 39-amino-linked adapters to the 39ends of viral dsRNAs followed by PCR amplification using a complementary primer. To investigate the possibility of endogenization of mitoviral sequences into the host genome (Liu et al., 2010; Chiba et al., 2011), we used three different mitovirus primer pairs and total cellular DNA from the host fungus as a template for PCR. The presence of mitochondrial DNA was verified with PCR amplification using the primer pair ML1/ ML2, which is specific for mitochondrial large subunit rDNA (White et al., 1990). Quantification of viral dsRNAs and transcripts with reverse transcription-quantitative PCR (RT-qPCR). For total RNA extrac-

tion, fungi were grown on cellophane-membrane-covered modified orange serum agar plates for 3–5 days. Total RNA was isolated from homogenized mycelia using TRI Reagent. cDNA was produced from 2 mg total RNA using the RevertAid First Strand cDNA Synthesis kit (Thermo Scientific) and random hexamer primers (Thermo Scientific). dsRNA was isolated using cellulose affinity chromatography from fungi grown for 14–20 days on cellophane-membrane-covered modified orange serum agar plates as described by Jurvansuu et al. (2014). Annealing of the adaptor oligonucleotide (59-TTGTATCGAGGCTGGTAAGC-39) and the 59-phosphorylated oligonucleotide (59-GCTTACCAGCCTCGATACAA-39) (100 mM each) was performed in 10 mM Tris (pH 8), 50 mM NaCl and 1 mM EDTA

Downloaded from www.microbiologyresearch.org by IP: 37.44.207.73 On: Sat, 28 Jan 2017 02:49:43

Journal of General Virology 96

Deep sequencing of mycoviral small RNAs at 95 uC for 5 min, followed by slow cooling to room temperature. T4 DNA ligase (Promega) was used to ligate annealed dsDNA adaptors to the isolated dsRNA. Ligation products were purified using the E.Z.N.A. Cycle Pure kit (Omega bio-tek). The Maxima H Minus double-stranded cDNA synthesis kit (Thermo Scientific) was used for synthesis of double-stranded cDNA using the oligonucleotide 59TTGTATCGAGGCTGGTAAGC-39. The reactions were treated with RNase I, cleaned with reagents from the E.Z.N.A. Cycle Pure kit and used for qPCR as described below. The cDNAs were diluted with an equivalent amount of nuclease-free water before amplification with qPCR using EvaGreen enzyme (Solis BioDyne) in a Rotor-GeneQ machine (Qiagen). The following qPCR primers were used: GAPDH (Lehr et al., 2009), HetPV13-an1 RdRp forward (59-GGTCAGGCTTCGATCTACGA-39) and reverse (59-AATTTGGTTGGCATGTAGCC-39), HetPV13-an1 CP forward (59-CAACCCAACTTGGCAGACTT-39) and reverse (59-GGTCGATCCAACAGGAGAAA-39), HetMV1-an1 RdRp forward (59-ATTACTCGGTTGTTGGCTCA39) and reverse (59-TCGTCCATGCTTTCCCAATA-39), HetRV6-pa19-c RdRp forward (59-ACATGCCCGCAGTCCGCTG-39) and reverse (59CGAGGTCATCGATAGTAGGGAATG-39) and HetPV7-pa1 RdRp forward (59-TGATTGACTGGTCTCGCTTC-39) and reverse (59GCAGATTTATGACGGGAGGT-39). qPCR standards for each primer pair were obtained by cloning the respective PCR product into the vector TOPO-pCR2.1, using the TOPO TA Cloning kit (Invitrogen), and preparing dilutions of the plasmid preparations containing 500 to 50 000 000 copies of the insert per 1 ml. All qPCR analyses were performed in triplicate as described by Jurvansuu et al. (2014). Bioinformatics. The small RNA contigs were assembled using the

Velvet sequence assembler program (Zerbino & Birney, 2008) using k-mer values ranging from 13 to 25. No contigs were obtained using k-mer values smaller or larger than this. Host rRNA sequences (18S rRNA, ITS1, 5.8S rRNA, ITS2 and 28S rRNA) were removed from the dataset using Novoalign (Novocraft Technologies, http://novocraft. com) supplemented with the specific script ‘Coderestar’ (Mirella Flores, International Potato Center, Peru, personal communication). After this, sequence assembly was conducted with Velvet and the resulting contigs were post-assembled into larger contigs using the AssemblyAssembler script (https://github.com/dzerbino/velvet/tree/ master/contrib/AssemblyAssembler1.3). The contigs produced with Velvet using k-mer values of 13-25 were compared with the nucleotide and protein sequences available in GenBank using BLASTN and BLASTX, respectively. Viral contigs were identified using the GenBank virus database (taxid: 10239). Sequence-guided assembly of siRNAs was conducted using the MAQ program (Li et al., 2008). The percentage coverage of small RNAs mapped to reference genome segments (referred to as sequence coverage) was determined using alignments between the original reference sequences and the MAQ output sequences indicating nucleotide sites covered by siRNAs. Sequence alignments and phylogenetic analyses were conducted using Geneious Pro 5.5.8 (Biomatters) and MEGA 5 (Tamura et al., 2011). An amino acid sequence alignment including sequences of HetMV1an1 and related mitovirus species was generated using MAFFT version 7.017. The resulting alignment was examined by eye according to the conserved motifs of mitoviral polymerases (Hong et al., 1999), which were accurately aligned. Dendrogram construction was conducted with Bayesian clustering using the MrBayes program implemented in Geneious Pro 5.5.8 and neighbour-joining clustering using MEGA 5. The Saccharomyces 23S RNA narnavirus was used as an outgroup. The Bayesian analyses were conducted with the gamma among-site rate variation and 1.16106 cycles for the MCMC algorithm of the MrBayes program, sampling one tree per 200 cycles and discarding 105 samples as burn-in. The Blosum matrix was selected as the protein http://vir.sgmjournals.org

evolution model based on the lowest Bayesian Information Criterion score obtained with ProtTest (Abascal et al., 2005).

ACKNOWLEDGEMENTS We are grateful to Juha Puranen, Tiia Lehmusmetsa¨ and Aazam Jalaledin for technical assistance, to Johanna Santala and Harri Vainio for help with data processing and to Muhammad Kashif for providing cDNA for mitovirus PCR amplification. This study was supported by the Academy of Finland (decision numbers 251193 and 258520) and the Finnish Forest Research Institute.

REFERENCES Abascal, F., Zardoya, R. & Posada, D. (2005). ProtTest: selection of

best-fit models of protein evolution. Bioinformatics 21, 2104–2105. Al Rwahnih, M., Daubert, S., Golino, D. & Rowhani, A. (2009). Deep

sequencing analysis of RNAs from a grapevine showing Syrah decline symptoms reveals a multiple virus infection that includes a novel virus. Virology 387, 395–401. Al Rwahnih, M., Daubert, S., Urbez-Torres, J. R., Cordero, F. & Rowhani, A. (2011). Deep sequencing evidence from single grapevine

plants reveals a virome dominated by mycoviruses. Arch Virol 156, 397–403. Bi, Y., Tugume, A. K. & Valkonen, J. P. T. (2012). Small-RNA deep

sequencing reveals Arctium tomentosum as a natural host of Alstroemeria virus X and a new putative Emaravirus. PLoS ONE 7, e42758. Bologna, N. G. & Voinnet, O. (2014). The diversity, biogenesis, and

activities of endogenous silencing small RNAs in Arabidopsis. Annu Rev Plant Biol 65, 473–503. Chang, S. S., Zhang, Z. & Liu, Y. (2012). RNA interference pathways

in fungi: mechanisms and functions. Annu Rev Microbiol 66, 305–323. Chiba, S., Salaipeth, L., Lin, Y.-H., Sasaki, A., Kanematsu, S. & Suzuki, N. (2009). A novel bipartite double-stranded RNA mycovirus

from the white root rot fungus Rosellinia necatrix: molecular and biological characterization, taxonomic considerations, and potential for biological control. J Virol 83, 12801–12812. Chiba, S., Kondo, H., Tani, A., Saisho, D., Sakamoto, W., Kanematsu, S. & Suzuki, N. (2011). Widespread endogenization of genome

sequences of non-retroviral RNA viruses into plant genomes. PLoS Pathog 7, e1002146. Chiba, S., Lin, Y. H., Kondo, H., Kanematsu, S. & Suzuki, N. (2013a).

Effects of defective interfering RNA on symptom induction by, and replication of, a novel partitivirus from a phytopathogenic fungus, Rosellinia necatrix. J Virol 87, 2330–2341. Chiba, S., Lin, Y. H., Kondo, H., Kanematsu, S. & Suzuki, N. (2013b).

A novel victorivirus from a phytopathogenic fungus, Rosellinia necatrix, is infectious as particles and targeted by RNA silencing. J Virol 87, 6727–6738. Coetzee, B., Freeborough, M. J., Maree, H. J., Celton, J. M., Rees, D. J. G. & Burger, J. T. (2010). Deep sequencing analysis of viruses infecting

grapevines: virome of a vineyard. Virology 400, 157–163. Drinnenberg, I. A., Weinberg, D. E., Xie, K. T., Mower, J. P., Wolfe, K. H., Fink, G. R. & Bartel, D. P. (2009). RNAi in budding yeast. Science

326, 544–550. Drinnenberg, I. A., Fink, G. R. & Bartel, D. P. (2011). Compatibility

with killer explains the rise of RNAi-deficient fungi. Science 333, 1592. Feldman, T. S., Morsy, M. R. & Roossinck, M. J. (2012). Are

communities of microbial symbionts more diverse than communities of macrobial hosts? Fungal Biol 116, 465–477.

Downloaded from www.microbiologyresearch.org by IP: 37.44.207.73 On: Sat, 28 Jan 2017 02:49:43

723

E. J. Vainio and others

Hammond, T. M., Andrewski, M. D., Roossinck, M. J. & Keller, N. P. (2008). Aspergillus mycoviruses are targets and suppressors of RNA

Lin, Y. H., Chiba, S., Tani, A., Kondo, H., Sasaki, A., Kanematsu, S. & Suzuki, N. (2012). A novel quadripartite dsRNA virus isolated from a

silencing. Eukaryot Cell 7, 350–357.

phytopathogenic filamentous fungus, Rosellinia necatrix. Virology 426, 42–50.

Heinze, C. (2012). A novel mycovirus from Clitocybe odora. Arch

Virol 157, 1831–1834. Hillman, B. I. & Cai, G. (2013). The family Narnaviridae: simplest of

RNA viruses. Adv Virus Res 86, 149–176. Hong, Y., Dover, S. L., Cole, T. E., Brasier, C. M. & Buck, K. W. (1999).

Multiple mitochondrial viruses in an isolate of the Dutch elm disease fungus Ophiostoma novo-ulmi. Virology 258, 118–127. Hu, Y., Stenlid, J., Elfstrand, M. & Olson, A˚. (2013). Evolution of

RNA interference proteins Dicer and Argonaute in Basidiomycota. Mycologia 105, 1489–1498. Hyder, R., Pennanen, T., Hamberg, L., Vainio, E. J., Piri, T. & Hantula, J. (2013). Two viruses of Heterobasidion confer beneficial, cryptic or

detrimental effects to their hosts in different situations. Fungal Ecol 6, 387–396. Ihrmark, K. (2001). Double-stranded RNA elements in the root rot

Liu, H., Fu, Y., Jiang, D., Li, G., Xie, J., Cheng, J., Peng, Y., Ghabrial, S. A. & Yi, X. (2010). Widespread horizontal gene transfer from

double-stranded RNA viruses to eukaryotic nuclear genomes. J Virol 84, 11876–11887. Ma´rquez, L. M., Redman, R. S., Rodriguez, R. J. & Roossinck, M. J. (2007). A virus in a fungus in a plant: three-way symbiosis required

for thermal tolerance. Science 315, 513–515. Martı´nez-A´lvarez, P., Vainio, E. J., Botella, L., Hantula, J. & Diez, J. J. (2014). Three mitovirus strains infecting a single isolate of Fusarium

circinatum are the first putative members of the family Narnaviridae detected in a fungus of the genus Fusarium. Arch Virol 159, 2153– 2155. Meister, G. & Tuschl, T. (2004). Mechanisms of gene silencing by

double-stranded RNA. Nature 431, 343–349.

fungus Heterobasidion annosum. PhD dissertation, Department of Forest Mycology and Pathology, Swedish University of Agricultural Sciences, Uppsala. ISBN 91–576–6094–8.

Morris, T. J. & Dodds, J. A. (1979). Isolation and analysis of double-

Ihrmark, K., Johannesson, H., Stenstro¨m, E. & Stenlid, J. (2002).

Mukasa, S. B., Rubaihayo, P. R. & Valkonen, J. P. T. (2006).

stranded RNA from virus-infected plant and fungal tissue. Phytopathology 69, 854–858.

Transmission of double-stranded RNA in Heterobasidion annosum. Fungal Genet Biol 36, 147–154.

Interactions between a crinivirus, an ipomovirus and a potyvirus in coinfected sweetpotato plants. Plant Pathol 55, 458–467.

Ihrmark, K., Stenstro¨m, E. & Stenlid, J. (2004). Double-stranded RNA

Nibert, M. L., Tang, J., Xie, J., Collier, A. M., Ghabrial, S. A., Baker, T. S. & Tao, Y. J. (2013). 3D structures of fungal partitiviruses. Adv Virus

transmission through basidiospores of Heterobasidion annosum. Mycol Res 108, 149–153.

Res 86, 59–85.

Jurvansuu, J., Kashif, M., Vaario, L., Vainio, E. J. & Hantula, J. (2014).

Nibert, M. L., Ghabrial, S. A., Maiss, E., Lesker, T., Vainio, E. J., Jiang, D. & Suzuki, N. (2014). Taxonomic reorganization of family

Partitiviruses of a fungal forest pathogen have species-specific quantities of genome segments and transcripts. Virology 462-463, 25–33. Kashif, M., Pietila¨, S., Artola, K., Jones, R. A. C., Tugume, A. K., Ma¨kinen, V. & Valkonen, J. P. T. (2012). Detection of viruses in

sweetpotato from Honduras and Guatemala augmented by deepsequencing of small-RNAs. Plant Dis 96, 1430–1437. Kashif, M., Hyder, R., De Vega Perez, D., Hantula, J. & Vainio, E. J. (2015). Heterobasidion wood decay fungi host diverse and globally

distributed viruses related to Helicobasidium mompa partitivirus V70. Virus Res 195 119–123. King, A. M., Adams, M. J., Lefkowitz, E. J. & Carstens, E. B. (2012).

Virus Taxonomy: IXth Report of the International Committee on Taxonomy of Viruses, vol. 9. San Diego, CA: Elsevier Academic Press. Kreuze, J. F., Perez, A., Untiveros, M., Quispe, D., Fuentes, S., Barker, I. & Simon, R. (2009). Complete viral genome sequence and

Partitiviridae and other recent progress in partitivirus research. Virus Res 188, 128–141. Olson, A˚., Aerts, A., Asiegbu, F., Belbahri, L., Bouzid, O., Broberg, A., Canba¨ck, B., Coutinho, P. M., Cullen, D. & other authors (2012).

Insight into trade-off between wood decay and parasitism from the genome of a fungal forest pathogen. New Phytol 194, 1001– 1013. Osaki, H., Nakamura, H., Sasaki, A., Matsumoto, N. & Yoshida, K. (2006). An endornavirus from a hypovirulent strain of the violet root

rot fungus, Helicobasidium mompa. Virus Res 118, 143–149. Pantaleo, V., Saldarelli, P., Miozzi, L., Giampetruzzi, A., Gisel, A., Moxon, S., Dalmay, T., Bisztray, G. & Burgyan, J. (2010). Deep

sequencing analysis of viral short RNAs from an infected Pinot Noir grapevine. Virology 408, 49–56.

discovery of novel viruses by deep sequencing of small RNAs: a generic method for diagnosis, discovery and sequencing of viruses. Virology 388, 1–7.

Poolpol, P. & Inouye, T. (1986). Enhancement of cucumber mosaic

Lakshman, D. K., Jian, J. & Tavantzis, S. M. (1998). A double-stranded

Romano, N. & Macino, G. (1992). Quelling: transient inactivation of

RNA element from a hypovirulent strain of Rhizoctonia solani occurs in DNA form and is genetically related to the pentafunctional AROM protein of the shikimate pathway. Proc Natl Acad Sci U S A 95, 6425– 6429. Lambden, P. R., Cooke, S. J., Caul, E. O. & Clarke, I. N. (1992).

Cloning of noncultivatable human rotavirus by single primer amplification. J Virol 66, 1817–1822. Lehr, N. A., Adomas, A., Asiegbu, F. O., Hampp, R. & Tarkka, M. T. (2009). WS-5995 B, an antifungal agent inducing differential gene

expression in the conifer pathogen Heterobasidion annosum but not in Heterobasidion abietinum. Appl Microbiol Biotechnol 85, 347–358.

virus multiplication by zucchini yellow mosaic virus in doubly infected cucumber plants. Ann Phytopathol Soc Jpn 52, 22–30. gene expression in Neurospora crassa by transformation with homologous sequences. Mol Microbiol 6, 3343–3353. Schoebel, C. N., Zoller, S. & Rigling, D. (2014). Detection and genetic characterisation of a novel mycovirus in Hymenoscyphus fraxineus, the causal agent of ash dieback. Infect Genet Evol 28, 78–86. Segers, G. C., Zhang, X., Deng, F., Sun, Q. & Nuss, D. L. (2007).

Evidence that RNA silencing functions as an antiviral defense mechanism in fungi. Proc Natl Acad Sci U S A 104, 12902–12906. Stenlid, J. & Redfern, D. B. (1998). Spread within the tree and stand.

Li, H., Ruan, J. & Durbin, R. (2008). Mapping short DNA sequencing

In Heterobasidion annosum: Biology, Ecology, Impact and Control, pp. 125–141. Edited by S. Woodward, J. Stenlid, R. Karjalainen & A. Hu¨ttermann. Wallingford, UK: CAB International.

reads and calling variants using mapping quality scores. Genome Res 18, 1851–1858.

Tamura, K., Peterson, D., Peterson, N., Stecher, G., Nei, M. & Kumar, S. (2011). MEGA5: molecular evolutionary genetics analysis using

724

Downloaded from www.microbiologyresearch.org by IP: 37.44.207.73 On: Sat, 28 Jan 2017 02:49:43

Journal of General Virology 96

Deep sequencing of mycoviral small RNAs maximum likelihood, evolutionary distance, and maximum parsimony methods. Mol Biol Evol 28, 2731–2739.

Vainio, E. J., Mu¨ller, M. M., Korhonen, K., Piri, T. & Hantula, J. (2014).

Tijsterman, M., Ketting, R. F. & Plasterk, R. H. (2002). The genetics of

Viruses accumulate in aging infection centers of a fungal forest pathogen. ISME J doi:10.1038/ismej.2014.145 [Epub ahead of print].

RNA silencing. Annu Rev Genet 36, 489–519.

White, T. J., Bruns, T., Lee, S. & Taylor, J. W. (1990). Amplification

Tuomivirta, T. T. & Hantula, J. (2003). Gremmeniella abietina

mitochondrial RNA virus S1 is phylogenetically related to the members of the genus Mitovirus. Arch Virol 148, 2429–2436. Tuomivirta, T. T., Uotila, A. & Hantula, J. (2002). Two independent double-stranded RNA patterns occur in the Finnish Gremmeniella abietina var. abietina type A. Forest Pathol 32, 197–205. Vainio, E. J., Korhonen, K., Tuomivirta, T. T. & Hantula, J. (2010). A

and direct sequencing of fungal ribosomal RNA genes for phylogenetics. In PCR Protocols – a Guide to Methods and Applications, pp. 315–322. Edited by M. Innis, D. H. Gelfand, J. J. Sninsky & T. J. White. San Diego, CA: Academic Press. Wu, Q., Luo, Y., Lu, R., Lau, N., Lai, E. C., Li, W. X. & Ding, S. W. (2010).

Virus discovery by deep sequencing and assembly of virus-derived small silencing RNAs. Proc Natl Acad Sci U S A 107, 1606–1611.

novel putative partitivirus of the saprotrophic fungus Heterobasidion ecrustosum infects pathogenic species of the Heterobasidion annosum complex. Fungal Biol 114, 955–965.

Wylie, S. J. & Jones, M. G. K. (2011). The complete genome sequence of a Passion fruit woodiness virus isolate from Australia determined using deep sequencing, and its relationship to other potyviruses. Arch Virol 156, 479–482.

Vainio, E. J., Kerio¨, S. & Hantula, J. (2011a). Description of a new

Wylie, S. J., Li, H., Dixon, K. W., Richards, H. & Jones, M. G. K. (2013).

putative virus infecting the conifer pathogenic fungus Heterobasidion parviporum with resemblance to Heterobasidion annosum P-type partitivirus. Arch Virol 156, 79–86. Vainio, E. J., Hakanpa¨a¨, J., Dai, Y.-C., Hansen, E., Korhonen, K. & Hantula, J. (2011b). Species of Heterobasidion host a diverse pool of

partitiviruses with global distribution and interspecies transmission. Fungal Biol 115, 1234–1243. Vainio, E. J., Hyder, R., Aday, G., Hansen, E., Piri, T., Dog˘mus¸Lehtija¨rvi, T., Lehtija¨rvi, A., Korhonen, K. & Hantula, J. (2012).

Exotic and indigenous viruses infect wild populations and captive collections of temperate terrestrial orchids (Diuris species) in Australia. Virus Res 171, 22–32. Yaegashi, H., Yoshikawa, N., Ito, T. & Kanematsu, S. (2013). A

mycoreovirus suppresses RNA silencing in the white root rot fungus, Rosellinia necatrix. Virology 444, 409–416. Yu, X., Li, B., Fu, Y., Jiang, D., Ghabrial, S. A., Li, G., Peng, Y., Xie, J., Cheng, J. & other authors (2010). A geminivirus-related DNA

mycovirus that confers hypovirulence to a plant pathogenic fungus. Proc Natl Acad Sci U S A 107, 8387–8392.

Population structure of a novel putative mycovirus infecting the conifer root-rot fungus Heterobasidion annosum sensu lato. Virology 422, 366–376.

Zerbino, D. R. & Birney, E. (2008). Velvet: algorithms for de novo short read assembly using de Bruijn graphs. Genome Res 18, 821–829.

Vainio, E. J., Piri, T. & Hantula, J. (2013a). Virus community dynamics

Zhang, X., Segers, G. C., Sun, Q., Deng, F. & Nuss, D. L. (2008).

in the conifer pathogenic fungus Heterobasidion parviporum following an artificial introduction of a partitivirus. Microb Ecol 65, 28–38. Vainio, E. J., Capretti, P., Motta, E. & Hantula, J. (2013b). Molecular

characterization of HetRV8-ir1, a partitivirus of the invasive conifer pathogenic fungus Heterobasidion irregulare. Arch Virol 158, 1613– 1615.

http://vir.sgmjournals.org

Characterization of hypovirus-derived small RNAs generated in the chestnut blight fungus by an inducible DCL-2-dependent pathway. J Virol 82, 2613–2619. Zhang, X., Shi, D. & Nuss, D. L. (2012). Variations in hypovirus

interactions with the fungal-host RNA-silencing antiviral-defense response. J Virol 86, 12933–12939.

Downloaded from www.microbiologyresearch.org by IP: 37.44.207.73 On: Sat, 28 Jan 2017 02:49:43

725