The virome from a collection of endomycorrhizal fungi reveals new viral taxa with unprecedented genome organization

Abstract

Mutualistic plant-associated fungi are recognized as important drivers in plant evolution, diversity, and health. The discovery that mycoviruses can take part and play important roles in symbiotic tripartite interactions has prompted us to study the viromes associated with a collection of ericoid and orchid mycorrhizal (ERM and ORM, respectively) fungi. Our study, based on high-throughput sequencing of transcriptomes (RNAseq) from fungal isolates grown in axenic cultures, revealed in both ERM and ORM fungi the presence of new mycoviruses closely related to already classified virus taxa, but also new viruses that expand the boundaries of characterized RNA virus diversity to previously undescribed evolutionary trajectories. In ERM fungi, we provide first evidence of a bipartite virus, distantly related to narnaviruses, that splits the RNA-dependent RNA polymerase (RdRP) palm domain into two distinct proteins, encoded by each of the two segments. Furthermore, in one isolate of the ORM fungus Tulasnella spp. we detected a 12 kb genomic fragment coding for an RdRP with features of bunyavirus-like RdRPs. However, this 12 kb genomic RNA has the unique features, for Bunyavirales members, of being tri-cistronic and carrying ORFs for the putative RdRP and putative nucleocapsid in ambisense orientation on the same genomic RNA. Finally, a number of ORM fungal isolates harbored a group of ambisense bicistronic viruses with a genomic size of around 5 kb, where we could identify a putative RdRP palm domain that has some features of plus strand RNA viruses; these new viruses may represent a new lineage in the Riboviria, as they could not be reliably assigned to any of the branches in the recently derived monophyletic tree that includes most viruses with an RNA genome.

Keywords: RNA-dependent RNA polymerase; ericoid mycorrhizal fungi; orchid mycorrhizal fungi; viruses.

Figure 1.

Main features of contigs related to members of Lenarviricota from ERM and ORM fungi. (A) Schematic representation of genome organization, with the main features. RdRP, RNA-dependent RNA polymerase; ORF, open reading frame; nt, nucleotides; A-sd, B-sd, C-sd, and D-sd are, respectively the A, B, C, and D subdomains of the RdRP palm domain. (B) Phylogenetic tree derived from alignments of the most closely related putative RdRP amino acid sequences inside the Lenarviricota clade. Model of substitution: Blosum62+F + I+G4. Consensus tree is constructed from 1,000 bootstrap trees. Log-likelihood of consensus tree: −49081.578022. At nodes, the percentage bootstrap values. The main existing and proposed taxonomic clades (class and family for the five smaller parentheses, phylum for the large one) are grouped with parenthesis. (C) TAE 1.5 per cent agarose gel of qPCR products to show that there is no DNA template corresponding to transcripts of OmSPV1 RNA1 and OmSPV1 RNA2. As a control for amplification we included a fragment of the O.maius actin gene. Lanes are labeled with the template used for the qPCR reaction and include an infected O.maius isolate (E27) and an uninfected isolate (E34). (D) 5′ and 3′ UTR of the sequence alignment of OmSPV1 RNA1 and OmSPV1 RNA2 genomic fragments.

Figure 2.

Northern blot analysis of narna-like contigs DN37559 and DN43802 representing the two genome segments of the OmSPV1. (A) Schematic representation of the position of the run-off transcript probes with codes identifying their orientation. In black, sense-oriented transcripts that hybridize with minus sense anti-genomic RNA intermediate. In red, antisense-oriented transcripts that hybridize with plus sense genomic RNA. (B) Top panels, autoradiography exposed 2 h with samples of total RNA (circa 3 μg/gel lane). The RNAs in the panel on the far left were hybridized first with probe no. 4 and subsequently with a ToBRFV probe (S1) that can be used as standard for RNA size (6.3 Kb for the genomic RNA and 0.7 kb for subgenomic RNA2). Lower panel is a methylene blue-stained membrane to show different rRNA loadings. Sample nomenclatures include E27 as the infected O.maius isolate and isolates O4 and E34 as negative controls. Mock is RNA extracted from a mock inoculated tomato plant. A blue arrow in this panel points to the position of the OmSP1 genomic RNAs, whereas black arrows points to the genomic and subgenomic RNA2 of ToBRFV.

Figure 3.

Protein alignments of contigs DN37559 and DN43802 representing the two genome segments of the OmSPV1. (A) Five protein sequences retrieved from different previously published assemblies from fungal RNAseq projects with some homology to OmSPV1 RNA1. Only the carboxy terminal of the proteins is shown. (B) Proteins from the same libraries with homology to OmSPV1 RNA2 ORF-encoded protein. Only the amino terminal of the protein alignment is displayed. OmSPV1 RNA1 is marked with a red asterisk and OmSPV1 RNA2 is marked with a blue asterisk. In the alignment, eight conserved regions already characterized for viral RdRPs could be observed and are marked with red rectangles for OmSPV1 RNA1-related viruses and blue rectangles for OmSPV1 RNA2-related sequences. Conserved regions are named I–VIII following Koonin (1991); regions IV–VII are linked to motifs A–D from the RdRP palm subdomain, respectively. 20074 is the code for the fungal isolate MUT4935 isolated from P.oceanica. F2 and F4 are libraries from mixed fungal isolates from esca-infected grapevines, and Holo3 is a library that comprises a number of fungal isolates from H.polii.

Figure 4.

Main features of the TuBlV1. (A) The genome organization of TuBlV1 with the main genomic features (ORFs and domains). Black bi-directional arrows indicate the position of RT-PCR amplification products to cover the whole genome. Unidirectional black and red arrows represent the position and orientation of the run-off transcripts used as probes in Northern blot analysis. Nc, nucleocapsid; RdRP, RNA-dependent RNA polymerase; ORF, open reading frame; nt, nucleotide. (B) Maximum likelihood phylogenetic tree derived from RdRP alignment of TuBlV1 with a number of bunyavirales representative of the main families in the order, (and two rhabdovirus used as outgroup). Amino acid substitution model is VT+F + I+G4. Log-likelihood of the tree: −200570.9602. Bootstrap values in percentage are displayed at each node. The tree is unrooted.

Figure 5.

Evidence of a contiguous 12 kb genomic segment for TuBlV1. (A) Overlapping RT-PCR of segments spanning the whole TuBlV1 genome. Lanes 1–6 correspond to segments A, B, F, C, D, and E of Fig. 4. M is the 1 kb Ladder. The three stronger bands correspond to 1, 3, and 7 kb, respectively. (B) Northern blot analysis of total RNA extracts from ORM fungal isolates O4 and O7 positive for CeAmV1 and for TuBlV1, respectively. Specific probes for each of the two viruses were hybridized in succession, in order to derive the specificity of each probe. Here, the result after the second hybridization is displayed, which therefore shows all the bands hybridizing with both probes. Red arrows point to the two bands specific for the ambivirus. Dotted black arrows point to the position of cross reactivity with rRNA. The black arrow points at the position of the TuBlV1 virus-specific RNA bands; upper panels are autoradiography exposed for 7 days. Lower panel is rRNA stained with methylene blue (the membrane picture was stretched vertically). Probes used are specified in Figs 4 and 8; rRNA, ribosomal RNA.

Figure 6.

Genome organization of ambiviruses and their putative RdRP palm subdomains. (A) Schematic representation of the genome organization of TuAmV1 and TuAmV2. ORFs are represented by green (ORFB) and orange (ORFA) arrows. (B) Alignment of conserved motifs of canonical viral RdRPs and ORFA proteins of ambiviruses. RdRPs were retrieved from alignment by Gorbalenya et al. (2002) and aligned using Clustal Omega. Conserved domains A, B, and C were selected from the alignment and sequences were aligned again on Clustal adding the ORFA sequences of ambiviruses discovered in the present study. Results were displayed on MEGA6. Canonical motifs are surrounded by the red rectangles with conserved residues marked by the red arrows. The putative motifs and conserved amino acids of ambiviruses surrounded by the black squares and black arrows indicate conserved residues. The sequences used for the alignment are as follows: tobacco vein mottling virus (TVMV, 8247947); feline calicivirus F9 (FCVF9, 130538); infectious flacherie virus (InFV, 3025415); Drosophila C virus (DCV, 2388673); human poliovirus type 3 Leon strain (PV3L, 130503); rice tungro spherical virus (RTSV, 9627951); cowpea severe mosaic virus (CPSMV, 549316); TuAmV1 (MN793991); TuAmV2 (MN793992); TuAmV5 (MN793996); TuAmV3 (MN793994); CeAmV1 (MN793993); and TuAmV4 (MN793994).

Figure 7.

Northern blot analysis of total RNA extracted from ORM fungal strains harboring ambiviruses. A: ORM fungal strains O4, O7, and O10, of which the latter was infected by TuAmV1. The probes used for TuAmV1 are presented at the bottom of each panel. Film was exposed to the membrane for 24 h. The blue arrow points to the position of the single specific band hybridizing with the probe. Dotted black arrows point to unspecific hybridization to rRNAs. (B) ORM fungal strains O4 and O7 of which the former hosts CeAmV1. Film was exposed to the membrane for 12 h. The left end panel was hybridized first with probe B2 and subsequently with a ToBRFV probe (S1) that can be used as standard for RNA size (6.3 Kb for the genomic RNA and 0.7 kb for subgenomic RNA2). In this panel a black arrow points to the position of ToBRFV genomic RNA and subgenomic RNA2, a blue arrow to the putative CeAmV1 genomic RNA and the red arrow to the position of the putative CeAmV1 dimer. Dotted black arrows point to the position of cross hybridizing rRNA. In both (A) and (B), lower panels are methylene blue-stained membranes rRNA loadings. Mock is RNA extracted from a mock inoculated tomato plant. At the top of each panel is a schematic representation of the position of the run-off transcript probes with codes that identify their orientation. In black, sense-oriented transcripts that hybridize with minus sense anti-genomic RNA intermediate. In red, antisense-oriented transcripts that hybridize with plus sense genomic RNA.