Journal of General Virology

Genome characterization of Botrytis virus F, a flexuous rod-shaped mycovirus resembling plant 'potex-like' viruses

Robyn L. J. Howitt,^1,2 Ross E. Beever,² Michael N. Pearson¹ and Richard L. S. Forster^3,4

¹ School of Biological Sciences, The University of Auckland, Private Bag 92019, Auckland, New Zealand
² Landcare Research, Private Bag 92170, Auckland, New Zealand
³ HortResearch, Private Bag 92169, Auckland, New Zealand
⁴ Genesis Research & Development Corporation Ltd, PO Box 50, Auckland, New Zealand

This study reports the first sequence of a flexuous rod-shaped mycovirus and also the first molecular characterization of a virus that infects the plant-pathogenic fungus Botrytis cinerea. The mycovirus Botrytis virus F (BVF) contains an ssRNA genome of 6827 nucleotides and a poly(A) tract at or very near the 3´ terminus. Computer analysis of the genomic cDNA sequence of BVF revealed two potential open reading frames (ORFs) encoding proteins of 212 kDa (ORF1) and 32 kDa (ORF2). ORF1 showed significant sequence identity to the RNA-dependent RNA polymerase (RdRp)-containing proteins of plant 'tymo-' and 'potex-like' viruses. However, the ORF1 protein contained an opal putative readthrough codon between the helicase and RdRp regions, a feature not seen in this position in 'tymo-' and 'potex-like' replicases sequenced to date. ORF2 shared amino acid similarity with coat proteins of plant 'potex-like' viruses. Three untranslated regions were present in the genome, comprising a region of 63 nucleotides preceding the initiation codon of ORF1, a 93 nucleotide stretch between ORFs 1 and 2 and a 3´-terminal region of 70 nucleotides preceding the poly(A) tract. The nucleotide sequence of a putative defective RNA (D-RNA) of 829 nucleotides was also determined. The D-RNA contained one potential ORF comprising the N-terminal region of the replicase fused in-frame to the C-terminal region of the coat protein. It is proposed that the mycovirus BVF belongs to a new, as yet unassigned genus in the plant 'potex-like' virus group.

The majority of characterized mycoviruses possess dsRNA genomes, either encapsidated in isometric particles (Buck, 1986) or occurring as unencapsidated elements associated with membranous vesicles in the cytoplasm (Hansen et al., 1985) or with mitochondrial fractions (Hong et al., 1998; Lakshman et al., 1998). Attention to date has focused particularly on those instances where dsRNA elements appear to confer hypovirulence on their hosts. Consequently, the bulk of available mycoviral sequence data relate to dsRNA genomes (Buck, 1998), although ssRNA sequences for a bacilliform particle from the button mushroom Agaricus bisporus (Revill et al., 1994) and an isometric particle from the plant-pathogenic fungus Sclerophthora macrospora (Yokoi et al., 1999) have been reported. Other morphological types of mycovirus such as flexuous and straight rod-shaped particles have been observed in four of the major taxonomic divisions of fungi (Buck, 1986), but these viruses are less well characterized.

Botrytis cinerea Pers. [teleomorph Botryotinia fuckeliana (de Bary) Whetzel] is an important pathogen in temperate climates, affecting a large number of economically important vegetable, flower and fruit crops (Coley-Smith et al., 1980). dsRNA elements associated with isometric particles have been reported in B. cinerea (Howitt et al., 1995; Vilches & Castillo, 1997; Castro et al., 1999). Bacilliform and flexuous rod-shaped particles have also been observed in B. cinerea (Howitt et al., 1995). The flexuous rod-shaped particles were comparable in size and morphology to ssRNA plant 'potex-like' viruses.

We report here the genome sequence of one virus, named Botrytis virus flexuous (BVF), reflecting the flexuous nature of the virus particles. This represents the first sequence of a virus of this morphology from a fungal host. Comparisons with known sequences provide an insight into the relationship between this mycovirus and similar viruses from other kingdoms.

Origin and culture of fungal strain. B. cinerea isolate RH106-10 (ICMP no. 12259) was collected from a field-grown strawberry plant at Massey, Auckland, New Zealand, in 1994. Growth conditions for B. cinerea were as described previously (Howitt et al., 1995). Briefly, the fungus was stored as a conidial suspension at –80 °C in 15 % glycerol. Conidia for inoculation of liquid medium were harvested from cultures grown on malt extract agar under a diurnal light regime. Mycelium for virus purification was grown in shake culture on Vogel's medium N (Vogel, 1964) supplemented with 0.5 % sucrose and 0.5 % yeast extract. Flasks (2 litre) containing 500 ml medium were inoculated with conidia to 10⁵/ml and incubated for 3 days on an orbital shaker (110 r.p.m.) at 20–25 °C.

Partial purification of viral particles. Frozen mycelium (10 g) was ground to a fine powder with a mortar and pestle in the presence of liquid nitrogen and partly purified as described by Howitt et al. (1995). The homogenate was extracted with chloroform and subjected to high- and low-speed centrifugation. At the second ultracentrifugation step, resuspended pellets were layered onto a 20 % sucrose cushion and finally resuspended in 20 mM sodium phosphate buffer (pH 7.0).

Extracts were stained with 2 % potassium phosphotungstate (pH 4.0) and examined in a Philips CM12 transmission electron microscope at 80 kV. Measurements were calibrated by using catalase crystals (Agar Scientific, cat. no. 124).

RNA extraction from partly purified preparations. Aliquots of the partly purified preparations (100 µl) were made up to 120 µl with buffer comprising 40 mM Tris–HCl (pH 8.0) and 12.5 mM MgCl₂ and incubated for 10 min at 37 °C with 20 U RNase-free DNase I (Life Technologies). Following DNase treatment, 100 µl portions were made up to 150 µl with buffer containing 10 mM Tris–HCl (pH 8.0), 5 mM EDTA and 0.5 % SDS. Samples were then incubated with 1 µl proteinase K (Life Technologies) at 20 Anson units/mg for 15 min at 42 °C. RNA was extracted with equal volumes of 1:1 (v/v) phenol–chloroform and precipitated with 2.5 vol ethanol in the presence of 300 mM sodium acetate (pH 5.2). The RNA was resuspended in 20 µl sterile water and stored at –20 °C.

cDNA synthesis and cloning. Unless otherwise indicated, all protocols used for the manipulation of nucleic acids and bacterial strains were those of Sambrook et al. (1989). The presence of a poly(A) tract was determined by the use of Dynabeads Oligo(dT)₂₅ as described by the manufacturer (Dynal). cDNA was synthesized and cloned by using the SuperScript plasmid system for cDNA synthesis and plasmid cloning (Life Technologies). Briefly, this system used a modified oligo(dT) NotI primer–adapter to prime first-strand cDNA synthesis. After second-strand synthesis, a SalI adapter was ligated to blunt-ended dsDNA, followed by NotI digestion. Following size fractionation, the cDNA was ligated into the NotI/SalI sites of plasmid pSport1 (Life Technologies). The plasmid containing the cDNA inserts was introduced into competent Escherichia coli MC1022 or DH5 cells. Plasmid DNA prepared from the transformants by the alkaline lysis method was digested with MluI to release the cDNA inserts. Resulting clones were mapped by restriction enzyme digestion and Southern blot analysis.

PCR amplification. Synthetic primers, complementary to the ends of the clones, were designed and used to amplify intervening sequences by RT–PCR. PCR mixtures (50 µl) contained 100 ng template cDNA, 10 pmol of each primer, 200 µM dNTPs and 1.25 mM MgCl₂ with the recommended buffer (Life Technologies). Two units of AmpliTaq DNA polymerase (Perkin Elmer) was added to the amplification following a 'hot start' of 5 min at 96 °C and 2 min at 94 °C; followed by 30 cycles of 40 s at 94 °C, 40 s at 58 °C and 2 min at 72 °C; and a final 10 min extension at 72 °C. Amplifications were performed in a Perkin Elmer Cetus thermal cycler.

Determination of the 5´ end. The presumptive 5´ cDNA clone was labelled with [-³²P]dCTP by using the Rediprime system. It was then used as a probe to identify further clones specific to the 5´ region by colony hybridization. Inserts were sized by agarose gel electrophoresis and mapped by enzyme digests.

The extreme 5´ RNA sequence was obtained by using a modified version of the Capfinder PCR cDNA Library Construction kit (Clontech). cDNA was synthesized with 200 U Superscript II RT (Life Technologies) and an oligonucleotide primer complementary to a region near the 5´ end of the presumptive 5´ cDNA clones. Included in this reaction was the Capswitch oligonucleotide from the kit, which served as a short, extended template attaching to the 5´ cap of the genomic RNA. cDNA was amplified directly by PCR by using the synthetic primer to the 5´ clones and a 5´ PCR primer complementary to the Capswitch oligonucleotide (supplied with the kit). The amplification products were either sequenced directly or ligated into the pGEM-T cloning vector (Promega) prior to sequencing.

Sequencing and sequence analysis. cDNA clones were sequenced in both forward and reverse directions. Nucleotide sequences were obtained either by sequencing of clones and subclones or by generating nested deletions of the subcloned cDNAs by using the Erase-a-Base system (Promega) according to the manufacturer's instructions. Sequencing was performed either manually with the 7-deaza-dGTP sequencing kit with Sequenase version 2.0 (USB) or with an Applied Biosystems model 373A sequencer using the Taq dideoxy terminator cycle sequencing method. Universal forward and reverse primers were used in conjunction with synthetic primers based on portions of the virus that had been sequenced.

Sequences were assembled by using the GCG 8.1 or 9.0 software package (Genetics Computer Group, Madison, WI, USA; Devereux et al., 1984) and GenBank searches performed with FASTA (GCG) or BLAST programs. Sequence alignment was performed with the CLUSTAL W package (Thompson et al., 1994). Internal methyltransferase, helicase and RNA-dependent RNA polymerase (RdRp) regions were aligned independently of the entire replicase alignment. A pairwise distance method using the Dayhoff PAM distance matrix from the PROTDIST program was used to infer phylogenetic relationships. Unrooted neighbour-joining distance trees were constructed by using the NEIGHBOR program. The robustness of each phylogeny was assessed by implementing bootstrap analysis consisting of 100 replicates in the SEQBOOT program (Felsenstein, 1985). PROTDIST, NEIGHBOR, SEQBOOT and the image-rendering program DRAWGRAM were implemented in the PHYLIP 3.5 package (Felsenstein, 1995). Sequence identity and similarity analyses were carried out with BLAST2 and CLUSTAL W programs.

In addition, nucleotide sequences of the replicase genes of BVF and other viruses were gapped according to the CLUSTAL W gapped aligned amino acid sequences and analysed by using DISCALC/DIPLOMO (Weiller & Gibbs, 1995). Unrooted neighbour-joining trees were created by using the NJTree program from a distance matrix resulting from the pairwise comparisons of the first and second nucleotides of each codon, with gaps excluded and allowing a correction for composition.

Sequences. GenBank accession numbers and acronyms for sequences used in analyses are listed in Table 1. Particular sequences were selected for comparison based on their similarity to BVF following database searches and potato leafroll virus (PLRV), southern bean mosaic virus (SBMV), tobacco rattle virus (TRV) and tobacco vein mottling potyvirus (TVMV) were included as outgroups.

Table 1. Plant viruses used in sequence analyses

			Accession no.
Genus	Virus	Acronym	Replicase	Coat protein	Reference
Luteoviridae
Polerovirus	Potato leafroll virus	PLRV	AB001894	–	Ohshima et al. (1993)
'Potex-like'
Allexivirus	Garlic virus A	GarV-A	AB010300	AB010300	Sumi et al. (1993)
	Garlic virus X	GVX	U89243	U89243	Song et al. (1998)
	Shallot virus X	SHVX	M97264	M97264	Kanyuka et al. (1992)
Capillovirus	Apple stem grooving virus	ASGV	D14995	D14995	Yoshikawa et al. (1992)
	Cherry virus A	CVA	X82547	X82547	Jelkmann (1995)
Carlavirus	Potato virus M	PVM	D14449	D14449	Zavriev et al. (1991)
	Garlic latent virus	GLV	Z68502	Z68502	Unpublished
Foveavirus	Apple stem pitting virus	ASPV	D21829	D21829	Jelkmann (1994)
	Grapevine rupestris stem pitting-associated virus	GRSPaV	AF026278	AF026278	Unpublished
Potexvirus	Plantago asiatica mosaic virus	PlAMV	Q07518	Q07506	Solovyev et al. (1994)
	Strawberry mild yellow edge-associated virus	SMYEV	D12517	D12517	Jelkmann et al. (1992)
Trichovirus	Apple chlorotic leafspot virus	ACLSV	D14996	P27737	Sato et al. (1993); German et al. (1990)
	Grapevine berry inner necrosis virus	GINV	–	D88448	Yoshikawa et al. (1997)
	Potato virus T	PVT	–	D10172	Ochi et al. (1992)
Vitivirus	Grapevine virus A	GVA	X75433	X75433	Minafra et al. (1997)
	Grapevine virus B	GVB	X75448	X75448	Saldarelli et al. (1996)
	Grapevine virus D	GVD	–	Y07764	Abou-Ghanem et al. (1997)
Potyviridae
Potyvirus	Tobacco vein mottling virus	TVMV	–	X04083	Domier et al. (1986)
'Tymo-like'
Tymovirus	Kennedya yellow mosaic virus	KYMV	D00637	–	Ding et al. (1990 a)
	Turnip yellow mosaic virus	TYMV	X16378	–	Dreher & Bransom (1992)
Marafivirus	Oat blue dwarf virus	OBDV	U87832	–	Edwards et al. (1997)
Unassigned
Sobemovirus	Southern bean mosaic virus	SBMV	NC_001625	–	Wu et al. (1987)
Tobravirus	Tobacco rattle virus	TRV	AF034622	–	Unpublished

We have separated members of the 'tymo-like' lineage of supergroup 3 positive-strand RNA viruses (Koonin & Dolja, 1993) into two groups. The 'tymo-like' virus group, with isometric particles, includes the genera Tymovirus and Marafivirus. The 'potex-like' group, with flexuous rod-shaped particles, includes the genera Allexivirus, Capillovirus, Carlavirus, Foveavirus, Potexvirus, Trichovirus and Vitivirus.

Virus purification

Following partial purification, flexuous rod-shaped particles with a modal length of approximately 720 nm were observed (Fig. 1).

Isolation of virus-specific cDNA clones

RNA extracted from the partly purified preparations was confirmed to be single stranded and the presence of a poly(A) tract was determined by using oligo(dT) magnetic beads and cDNA priming with oligo(dT) primers. Labelled cDNA of the partly purified mycoviral RNA migrated as a single species close to potato virus X RNA (6435 bp), indicating a size of ~6500 bp (data not shown).

Fig. 1. Electron micrograph of flexuous rod-shaped particles in a partly purified preparation of Botrytis cinerea isolate RH106-10, negatively stained with 2 % potassium phosphotungstate, pH 4.0. Bar, 200 nm.

The cloning strategy for BVF RNA sequencing is outlined in Fig. 2. Three non-overlapping clones, BVF23 (1035 bp), BVF24 (1173 bp) and BVF37 (1976 bp), representing 60 % of the total genome, were produced by using the SuperScript cloning protocol. Clone BVF37 contained a poly(A) tract at the 3´ end and a GDD motif at the 5´ end, and BVF24 contained a GXXGXGKS/T motif. Conserved RdRp (POL or GDD) (Koonin, 1991) and helicase (HEL or GKS) (Gorbalenya & Koonin, 1989) sequence motifs, along with methyltransferase (MTR) motifs (Rozanov et al., 1992), are found in the replicase genes of many positive-strand RNA viruses.

Additional clones, BVF1 and BVF3, were obtained by RT–PCR primed with BVF-specific primers. Sequence analysis of clone BVF1 revealed a single NotI site, indicating that clones BVF23 and BVF24 represent a contiguous stretch of sequence. Clone BVF5 was identified by colony hybridization with a probe generated from BVF23 and clone BVF15 was obtained by using the Capfinder kit (Clontech). Clone BVF15 and two additional clones generated from independent PCRs were shown to be identical by sequence analysis.

Genome organization

The genome size of BVF RNA, excluding the poly(A) tract, was 6827 nucleotides (GenBank accession no. AF238884). Computer analysis of the nucleotide sequence revealed the presence of two putative open reading frames (ORFs) (Fig. 2). Three untranslated regions were present in the genome, comprising a region of 63 nucleotides preceding the initiation codon of ORF1, a 93 nucleotide stretch between ORFs 1 and 2 and a 3´-terminal region of 70 nucleotides followed by a poly(A) tract. The existence of a 5´ cap was deduced from the presence of a methyltransferase region in the replicase gene and from the successful use of a Capswitch oligonucleotide to derive the 5´-terminal region. The fact that two independent cDNA clones derived by this approach provided sequences identical to that of clone BVF15 suggests that the entire 5´ region has been identified correctly.

Fig. 2. Schematic representation of the genome organization and cloning strategy of BVF. (a) Positions of ORFs (boxed) and the size of their putative products are shown. The relative positions of the conserved domains of the replicase gene are indicated with shaded boxes: MTR, methyltransferase; HEL, helicase; POL, RdRp. The position of the putative readthrough codon is indicated with a diamond. Boxes with cross-hatching represent the portions of the 5´ and 3´ termini that are found in a putative D-RNA species. (b) D-RNA species that has an ORF with the potential to encode a replicase–coat protein fusion protein. (c) Map of overlapping cDNA clones. Unshaded boxes indicate clones obtained with specific primers.

A putative protein with predicted molecular mass 19.7 kDa (p20) overlapped ORF1 by 467 nucleotides. No AUG initiation codon was detected for this 176 amino acid sequence, which terminated with an opal (UGA) codon at nucleotide 530. This putative reading frame was in a –1 frame relative to the two other ORFs. The base content of the p20 coding region was 26.0 % A, 34.0 % C, 20.5 % U and 19.5 % G and encoded a proline-/serine-rich protein. Database searches with the nucleotide and amino acid sequences of p20 did not reveal any significant homology with other protein sequences.

The cDNA to a putative defective RNA (D-RNA) of 829 nucleotides was detected following colony hybridization with probe BVF5. The D-RNA consisted of a 5´ region of the parental genome (nucleotides 17–315) fused to the 3´ terminus (nucleotides 6298–6827), complete with poly(A) tract (Fig. 2). This D-RNA, encoding a putative protein with predicted molecular mass 24.9 kDa (p25), contained the first 84 amino acids of ORF1 and the last 152 amino acids of ORF2.

Table 2. Percentage amino acid identity and similarity between the replicase and coat protein 'salt bridge' regions of BVF and selected plant viruses

Percentage similarities are shown in parentheses. Columns (i) show comparisons between BVF and the first virus listed unless italicized, where comparison is with the second virus. Columns (ii) show intragenic comparisons between the pairs of viruses listed. See Table 1 for details of sequences. ND, No data available; –, alignment not attempted because sequences were too disparate or comparison was not relevant.

ORF1

The initiation codon for ORF1 was in a favourable context for translation in filamentous fungi, having a cytosine in the +5 position and an adenine in the –3 position (Ballance, 1990). This ORF, encoding a putative protein with expected molecular mass 153 kDa (p153), terminated at an opal (UGA) codon at nucleotide 4192. Readthrough of this in-frame stop codon would result in a larger protein with molecular mass 212 kDa (p212), terminating at an amber (UAG) codon at nucleotide 5752. The presence of the opal stop codon was confirmed by sequencing three independently derived clones. Database searches of the amino acid sequence of p212 showed sequence identity to the replicase genes of ssRNA plant 'tymo-' and 'potex-like' viruses, belonging to the Sindbis-like supergroup of positive-strand RNA viruses. Alignments were made of the entire replicase and the three internal conserved replicase regions containing the methyltransferase, helicase and RdRp domains. The boundaries of these internal regions were defined by Morozov et al. (1990) and include all motifs recognized by Koonin & Dolja (1993). When aligned with a selection of replicases from the 'tymo-' and 'potex-like' viruses, BVF p212 showed highest amino acid identity to replicases from allexivirus [garlic virus X (GVX), 23 %], tymovirus [turnip yellow mosaic virus (TYMV), 23 %], potexvirus [strawberry mild yellow edge-associated virus (SMYEV), 22 %] and vitivirus [grapevine virus B (GVB), 22 %] (Table 2). Alignments of the three internal, conserved replicase regions also showed high homology to those regions in 'tymo-' and 'potex-like' viruses. In the methyltransferase region, BVF showed highest amino acid identity to the 'tymo-like' viruses [TYMV, oat blue dwarf virus (OBDV), 34 %], but was closer to the 'potex-like' viruses in the helicase (GVX, SMYEV, 28 %) and RdRp [garlic virus A (GarV-A), 40 %; SMYEV, cherry virus A (CVA), 39 %] regions (Table 2). An alignment of conserved RdRp sequence motifs for BVF and selected plant viruses is shown in Fig. 3.

Fig. 3. Amino acid alignment of the putative RdRp domains of the replicase proteins of BVF and selected plant viruses. Motifs I–VIII are based on those of Koonin & Dolja (1993). Conserved residues in BVF and the most closely related plant viruses are shown by asterisks, similar residues by colons. Bold asterisks denote conservation across all viruses aligned, bold colons denote similarity across all viruses. Residues conserved in at least eight viruses are shown in bold. Numbers in brackets refer to the amino acid position in the ORF.

Phylogenetic analysis of this RdRp region revealed that BVF clustered amongst the 'potex-like' group, supported by robust bootstrap values (Fig. 4). The polerovirus PLRV and sobemovirus SBMV were included in this analysis as they share the highest amino acid identity to the RdRps of the other two ssRNA mycoviruses sequenced to date.

Unrooted neighbour-joining trees constructed from pairwise distances of the first and second nucleotides of each codon of the replicase genes, with a correction for composition (high cytosine bias), produced similar groupings to those presented in Table 2 for the entire replicase and the helicase and RdRp regions. However, in the methyltransferase region, BVF grouped closer to the 'potex-like' viruses than to the 'tymo-like' viruses (data not shown).

The base content of BVF ORF1 RNA was 24.4 % A, 32.1 % C, 22.7 % U and 20.8 % G. A similar nucleotide composition with high cytosine content is also observed in ORF1 of tymoviruses, the marafivirus OBDV and some potexviruses, including Plantago asiatica mosaic virus (PlAMV) and, to a lesser extent, foxtail mosaic virus.

Fig. 4. Phylogram of the putative RdRp domains of BVF and selected plant viruses. Numbers indicate the percentage of bootstrap replicates out of 100 that support each branch node (only values >50 % are shown). Refer to Table 1 for details of sequences. Bar, 10 substitutions per 100 amino acids.

ORF2

ORF2 (nucleotides 5848–6754) encodes a putative protein with expected molecular mass 32 kDa (p32). Again, the initiation codon was in a favourable context for translation. Database searches with the 302 amino acid sequence of p32 revealed homology to capillo- and trichoviruses. The BVF p32 protein contained conserved residues (Fig. 5) that have been associated with the putative salt bridge and corresponding hydrophobic core in flexuous rod-shaped ssRNA plant viruses (Dolja et al., 1991). In this salt bridge region, BVF showed the highest amino acid identity to the trichovirus apple chlorotic leafspot virus (ACLSV) (33 %) and the vitivirus GVB (29 %) (Table 2). A similar relationship, although less well supported, was observed following phylogenetic analysis of this region (Fig. 6). The putative coat protein of BVF was unusual in having a long C-terminal region when compared with the coat proteins of plant viruses in this alignment (Fig. 5). The base content of BVF ORF2 RNA (26.4 % A, 35.4 % C, 17.2 % U and 21.0 % G) was high in cytosine, more typical of a number of potexvirus coat protein genes than those of the vitiviruses.

Fig. 5. Amino acid alignment of segments of the coat proteins of BVF and flexuous rod-shaped plant viruses potentially involved in salt-bridge formation. The conserved positively and negatively charged residues (Arg and Asp, respectively), proposed to form a salt bridge crucial for protein structure, are shown in bold (modified from Dolja et al., 1991). Identical residues are indicated by asterisks and similar residues by colons. The distances from the protein termini and between the aligned segments are indicated by numbers in brackets.

Comparison of the full amino acid sequence of BVF showed that the putative BVF replicase gene is most closely related to replicase genes of both 'tymo-' and 'potex-like' viruses. Analysis of the three most conserved regions of the replicase showed features more closely resembling 'tymo-like' viruses in the N-terminal region, but more similar to 'potex-like' viruses in the C-terminal region. In the methyltransferase region, the BVF replicase showed significant amino acid sequence identity to similar regions in 'tymo-like' viruses. This is not surprising, considering that the first half of BVF and 'tymo-like' virus replicases are both cytosine-rich. Two alternative explanations can be suggested for BVF having a high cytosine content. The first would be that the cytosine-richness is an ancestral feature of BVF and does indeed reflect a close affinity with the 'tymo-like' viruses. The second would be that the overlapping reading frame at the N-terminal region of the BVF replicase, being proline-rich, has driven secondary cytosine-enrichment of this region of the BVF replicase, which is therefore not indicative of its ancestral composition (Gibbs & Keese, 1994). If the latter is the correct explanation, a comparison of pairwise distances with a correction for this high cytosine nucleotide bias would show the true taxonomic relationships of this region. In such a comparison, BVF groups closer to the methyltransferase regions of 'potex-like' viruses.

An unusual feature of BVF is the putative readthrough opal (UGA) codon in the replicase, confirmed by the sequencing of several independent clones. This phenomenon has not been reported in the replicases of other 'potex-like' viruses. In the tymovirus TYMV, readthrough of an amber (UAG) codon results in the extension of the 206 kDa protein to a 221 kDa protein (Bransom et al., 1995). However, in BVF, the type of stop codon (i.e. opal), its position in the replicase (positioned between the helicase and RdRp regions) and the immediate surrounding nucleotide sequences are more typical of readthrough codons found in RdRp-encoding RNA1 molecules of the bipartite genomes of ssRNA straight rod-shaped viruses. These include pea early browning virus (MacFarlane et al., 1989) and TRV (Hamilton et al., 1987) in the genus Tobravirus and soil-borne wheat mosaic virus in the genus Furovirus (Shirako & Wilson, 1993). These plant viruses and BVF share the sequence aUAGc(g/a)(g/a)t in this position. The cytosine residue located immediately downstream of the UAG codon has been shown to be important for efficient translational readthrough in mammalian viruses (Li & Rice, 1993).

We can therefore suggest a translation strategy for BVF based on examples from other viruses (Buck, 1996). The methyltransferase and helicase domains are located on a protein that is translated from the 5´ ORF and the RdRp domain is translated by readthrough of the stop codon to give a fusion protein. In these instances, the methyltransferase and helicase regions are present in excess over the RdRp domain and the downstream coat protein is translated from a subgenomic RNA. In vitro translation studies, coupled with the identification of an as yet undetected subgenomic RNA species in BVF, could confirm this strategy.

Fig. 6. Phylogram of the putative 'salt bridge' region corresponding to amino acids 117–138 of the BVF coat protein. Numbers indicate the percentage of bootstrap replicates out of 100 that support each branch node (only values >50 % are shown). Refer to Table 1 for details of sequences. Bar, 10 substitutions per 100 amino acids.

In the BVF coat protein, the regions of highest homology align with the most conserved central core regions of the coat proteins of plant 'potex-like' viruses. Although amino acid identity is highest with the genera Capillovirus, Trichovirus and Vitivirus, the BVF coat protein is still distinct enough to be grouped in its own genus. Phylogenetic analysis, although not robust, reveals BVF clustering with the 'capillo' rather than the 'potex/carla' lineage of coat proteins (Koonin & Dolja, 1993). It is interesting to note that there is moderately high amino acid identity (39 %) between BVF and the capillovirus CVA in the RdRp region at the C terminus of the replicase gene.

An interesting feature of BVF, when compared with other 'potex-' and 'tymo-like' viruses, is the apparent lack of a movement protein. Many plant viruses encode movement proteins that interact with plant host plasmodesmata to allow the passage of infectious viral material from cell to cell. Movement proteins have the capacity to move from cell to cell, dilate plasmodesmal microchannels (normally up to 4 nm in diameter; Fisher, 1999) and facilitate the passage of viral nucleic acids (Lucas & Gilbertson, 1994). The mycovirus BVF, having a coenocytic host, presumably does not require similar movement proteins. In B. cinerea, it is unlikely that the ~150–250 nm septal pore that separates individual 'cells' (Vilches & Castillo, 1997; Gull & Trinci, 1971) would be a barrier to the movement of mycovirus particles. However, the possible role of Woronin bodies (Gull, 1978) in the transport of mycoviruses remains unknown.

The putative 20 kDa protein that overlaps the replicase gene may be analogous to the overlapping movement proteins of the tymoviruses, although much smaller in size. Alternative initiation codons (AUU, CUG, UUG) are known to be used in plants (Gordon et al., 1992) and AUU was assigned as a possible initiation codon for ORFs in the allexivirus GVX (Song et al., 1998) and the potexvirus SMYEV (Jelkmann et al., 1992). In the BVF genome, an in-frame AUU codon is present only at nucleotide 2. Alternatively, an in-frame CUG codon is present at nucleotide 68. In the marafivirus OBDV, a similar-sized overlapping protein is present, also lacking an AUG initiation codon. OBDV resembles the tymoviruses closely in genome sequence, organization and expression strategy (Edwards et al., 1997), but lacks the large (49–70 kDa) overlapping movement proteins of the tymoviruses. OBDV is restricted to phloem tissues in its plant host (Edwards et al., 1997). The small overlapping protein of OBDV has significant amino acid identity to those of the tymoviruses. The putative overlapping protein, p20, of BVF has no such sequence similarity but does have a similar proline-rich composition. Unlike the two BVF ORFs, the putative reading frame of p20 has the highest cytosine content in the second rather than third codon position. Due to this nucleotide bias, it is possible that the original BVF genome comprised ORFs 1 and 2, the putative replicase and coat protein genes, and that the putative p20 reading frame arose later by overprinting (Gibbs & Keese, 1994).

Assuming that the putative D-RNA found in BVF preparations is not a cloning artefact, this RNA may have arisen during replication by an internal deletion event resulting in the fusion of the 5´ and 3´ ends of the parental genome. D-RNAs are found in animal, plant and fungal viruses (Roux, 1994). Deletion mutants derived from dsRNA genetic elements have been reported previously in mycoviruses. In Saccharomyces cerevisiae killer viruses, these mutants result in neither toxin production nor immunity (Nuss, 1988), and in Cryphonectria parasitica they are proposed to be responsible for the observed complex banding patterns of dsRNA (Shapira et al., 1991). It has yet to be determined whether this D-RNA is capable of replicating or if it has any modulating effect on BVF. Interestingly, two potexviruses, cassava common mosaic virus (Calvert et al., 1996) and clover yellow mosaic virus (White et al., 1992), also produce D-RNAs that invariably encode a single ORF involving an in-frame fusion of the N terminus of the replicase and the C terminus of the coat protein. White et al. (1992) suggested that this conservation of reading frame, also a feature of the BVF D-RNA, may be essential for RNA stability.

In summary, BVF has features typical of four groups of ssRNA plant viruses: the genera Furovirus and Tobravirus with straight rod-shaped particles, the 'tymo-like' viruses with isometric particles and the 'potex-like' viruses with flexuous rod-shaped particles. This is quite distinct from the other two ssRNA mycoviruses sequenced to date. The mushroom bacilliform virus and the Sclerophthora macrospora virus share sequence identity to luteo- and sobemoviruses (Yokoi et al., 1999). Similarities that the BVF genome shares with 'tymo-like' genomes are the size (212 kDa) and high cytosine content of the putative replicase gene. Differences are the lack of both a large overlapping movement protein at the 5´ end of the genome and a 16 nucleotide 'tymobox' sequence at the 3´ end of the replicase gene, a hallmark of the tymoviruses (Ding et al., 1990 b). Also, BVF contains coat protein sequence motifs typical of flexuous rod-shaped rather than isometric particles.

Amino acid sequence identities of the conserved helicase and RdRp replicase regions and the coat protein genes are greatest to those of 'potex-like' viruses. The major difference between BVF and these plant viruses is the lack of a movement protein, either as a single polypeptide (Capillovirus, Foveavirus, Trichovirus, Vitivirus) or as a triple gene block or similar structure (Allexivirus, Carlavirus, Potexvirus). Amino acid sequences of both the putative replicase and coat protein genes, along with conserved motifs present in the coat protein gene, support the classification of BVF in the plant virus 'potex-like' group. However, when a comparison is made of amino acid identities for these genes between existing 'potex-like' genera (Table 2), it is apparent that the mycovirus BVF is distinct enough to belong to a new genus in this group.

The authors wish to thank Professor Adrian Gibbs and Dr Mark Gibbs (Bioinformatics Lab, School of Biological Sciences, ANU, Canberra, Australia) for advice and assistance with sequence analysis and Dr David Beck (HortResearch, Auckland, New Zealand) for helpful discussion. The senior author acknowledges the provision of laboratory facilities by HortResearch, Auckland.

The GenBank accession number of the sequence reported in this paper is AF238884.

Abou-Ghanem, N., Saldarelli, P., Minafra, A., Buzkan, N., Castellano, M. A. & Martelli, G. P. (1997). Properties of grapevine virus D, a novel putative trichovirus. Journal of Plant Pathology 78, 15–25.

Ballance, D. J. (1990). Transformation systems for filamentous fungi and an overview of fungal gene structure. In Molecular Industrial Mycology: Systems and Applications for Filamentous Fungi, vol. 8, pp. 1–29. Edited by S. A. Long & R. M. Berka. New York: Dekker.

Bransom, K. L., Weiland, J. J., Tsai, C.-H. & Dreher, T. W. (1995). Coding density of the turnip yellow mosaic virus genome: roles of the overlapping coat protein and p206-readthrough coding regions. Virology 206, 403–412.

Buck, K. W. (1986). Fungal virology – an overview. In Fungal Virology, pp. 1–84. Edited by K. W. Buck. Boca Raton, FL: CRC Press.

Buck, K. W. (1996). Comparison of the replication of positive-stranded RNA viruses of plants and animals. Advances in Virus Research 47, 159–251.

Buck, K. W. (1998). Molecular variability of viruses of fungi. In Molecular Variability of Fungal Pathogens, pp. 53–72. Edited by P. D. Bridge, Y. Couteaudier & J. M. Clarkson. Wallingford, UK: CAB International.

Calvert, L. A., Cuervo, M. I., Ospina, M. D., Fauquet, C. M. & Ramirez, B.-C. (1996). Characterization of cassava common mosaic virus and a defective RNA species. Journal of General Virology 77, 525–530.

Castro, M., Kramer, K., Valdivia, L., Ortiz, S., Benavente, J. & Castillo, A. (1999). A new double-stranded RNA mycovirus from Botrytis cinerea. FEMS Microbiology Letters 175, 95–99.

Coley-Smith, J. R., Verhoeff, K. & Jarvis, W. R. (1980). The Biology of Botrytis. London: Academic Press.

Devereux, J., Haeberli, P. & Smithies, O. (1984). A comprehensive set of sequence analysis programs for the VAX. Nucleic Acids Research 12, 387–395.

Ding, S., Keese, P. & Gibbs, A. (1990 a). The nucleotide sequence of the genomic RNA of kennedya yellow mosaic tymovirus-Jervis Bay isolate: relationships with potex- and carlaviruses. Journal of General Virology 71, 925–931.

Ding, S. W., Howe, J., Keese, P., Mackenzie, A., Meek, D., Osorio-Keese, M., Skotnicki, M., Srifah, P., Torronen, M. & Gibbs, A. (1990 b). The tymobox, a sequence shared by most tymoviruses: its use in molecular studies of tymoviruses. Nucleic Acids Research 18, 1181–1187.

Dolja, V. V., Boyko, V. P., Agranovsky, A. A. & Koonin, E. V. (1991). Phylogeny of capsid proteins of rod-shaped and filamentous RNA plant viruses: two families with distinct patterns of sequence and probably structure conservation. Virology 184, 79–86.

Domier, L. L., Franklin, K. M., Shahabuddin, M., Hellmann, G. M., Overmeyer, J. H., Hiremath, S. T., Siaw, M. F., Lomonossoff, G. P., Shaw, J. G. & Rhoads, R. E. (1986). The nucleotide sequence of tobacco vein mottling virus RNA. Nucleic Acids Research 14, 5417–5430.

Dreher, T. W. & Bransom, K. L. (1992). Genomic RNA sequence of turnip yellow mosaic virus isolate TYMC, a cDNA-based clone with verified infectivity. Plant Molecular Biology 18, 403–406.

Edwards, M. C., Zhang, Z. & Weiland, J. J. (1997). Oat blue dwarf marafivirus resembles the tymoviruses in sequence, genome organization, and expression strategy. Virology 232, 217–229.

Felsenstein, J. (1985). Confidence limits on phylogenies: an approach using the bootstrap. Evolution 39, 783–791.

Felsenstein, J. (1995). PHYLIP version 3.5. Department of Genetics, University of Washington, Seattle, WA, USA.

Fisher, D. B. (1999). The estimated pore diameter for plasmodesmal channels in the Abutilon nectary trichome should be about 4 nm, rather than 3 nm. Planta 208, 299–300.

German, S., Candresse, T., Lanneau, M., Huet, J. C., Pernollet, J. C. & Dunez, J. (1990). Nucleotide sequence and genomic organization of apple chlorotic leaf spot closterovirus. Virology 179, 104–112.

Gibbs, A. & Keese, P. K. (1994). In search of the origins of viral genes. In Molecular Basis of Viral Evolution, pp. 76–90. Edited by A. J. Gibbs, C. H. Calisher & F. Garcia-Arenal. Cambridge: Cambridge University Press.

Gorbalenya, A. E. & Koonin, E. V. (1989). Viral proteins containing the purine NTP-binding sequence pattern. Nucleic Acids Research 17, 8413–8440.

Gordon, K., Futterer, J. & Hohn, T. (1992). Efficient initiation of translation at non-AUG triplets in plant cells. Plant Journal 2, 809–813.

Gull, K. (1978). Form and function of septa in filamentous fungi. In The Filamentous Fungi, vol. 3, pp. 78–93. Edited by J. E. Smith & D. R. Berry. New York: John Wiley.

Gull, K. & Trinci, A. P. J. (1971). Fine structure of spore germination in Botrytis cinerea. Journal of General Microbiology 68, 207–220.

Hamilton, W. D. O., Boccara, M., Robinson, D. J. & Baulcombe, D. C. (1987). The complete nucleotide sequence of tobacco rattle virus RNA-1. Journal of General Virology 68, 2563–2575.

Hansen, D. R., Van Alfen, N. K., Gillies, K. & Powell, W. A. (1985). Naked dsRNA associated with hypovirulence of Endothia parasitica is packaged in fungal vesicles. Journal of General Virology 66, 2605–2614.

Hong, Y., Cole, T. E., Brasier, C. M. & Buck, K. W. (1998). Novel structures of two virus-like RNA elements from a diseased isolate of the Dutch elm disease fungus, Ophiostoma novo-ulmi. Virology 242, 80–89.

Howitt, R. L. J., Beever, R. E., Pearson, M. N. & Forster, R. L. S. (1995). Presence of double-stranded RNA and virus-like particles in Botrytis cinerea. Mycological Research 99, 1472–1478.

Jelkmann, W. (1994). Nucleotide sequences of apple stem pitting virus and of the coat protein gene of a similar virus from pear associated with vein yellows disease and their relationship with potex- and carlaviruses. Journal of General Virology 75, 1535–1542.

Jelkmann, W. (1995). Cherry virus A: cDNA cloning of dsRNA, nucleotide sequence analysis and serology reveal a new plant capillovirus in sweet cherry. Journal of General Virology 76, 2015–2024.

Jelkmann, W., Maiss, E. & Martin, R. R. (1992). The nucleotide sequence and genome organization of strawberry mild yellow edge-associated potexvirus. Journal of General Virology 73, 475–479.

Kanyuka, K. V., Vishnichenko, V. K., Levay, K. E., Kondrikov, D. Yu., Ryabov, E. V. & Zavriev, S. K. (1992). Nucleotide sequence of shallot virus X RNA reveals a 5´-proximal cistron closely related to those of potexviruses and a unique arrangement of the 3´-proximal cistrons. Journal of General Virology 73, 2553–2560.

Koonin, E. V. (1991). The phylogeny of RNA-dependent RNA polymerases of positive-strand RNA viruses. Journal of General Virology 72, 2197–2206.

Koonin, E. V. & Dolja, V. V. (1993). Evolution and taxonomy of positive-strand RNA viruses: implications of comparative analysis of amino acid sequences. Critical Reviews in Biochemistry and Molecular Biology 28, 375–430.

Lakshman, D. K., Jian, J. & Tavantzis, S. M. (1998). A double-stranded 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. Proceedings of the National Academy of Sciences, USA 95, 6425–6429.

Li, G. & Rice, C. M. (1993). The signal for translational readthrough of a UGA codon in Sindbis virus RNA involves a single cytidine residue immediately downstream of the termination codon. Journal of Virology 67, 5062–5067.

Lucas, W. J. & Gilbertson, R. L. (1994). Plasmodesmata in relation to viral movement within leaf tissues. Annual Review of Phytopathology 32, 387–411.

MacFarlane, S. A., Taylor, S. C., King, D. I., Hughes, G. & Davies, J. W. (1989). Pea early browning virus RNA1 encodes four polypeptides including a putative zinc-finger protein. Nucleic Acids Research 17, 2245–2260.

Minafra, A., Saldarelli, P. & Martelli, G. P. (1997). Grapevine virus A: nucleotide sequence, genome organization, and relationship in the Trichovirus genus. Archives of Virology 142, 417–423.

Morozov, S. Yu., Kanyuka, K. V., Levay, K. E. & Zavriev, S. K. (1990). The putative RNA replicase of potato virus M: obvious sequence similarity with potex- and tymoviruses. Virology 179, 911–914.

Nuss, D. L. (1988). Deletion mutants of double-stranded RNA genetic elements found in plants and fungi. In RNA Genetics, vol. 2, pp. 187–210. Edited by E. Domingo, J. J. Holland & P. Ahlquist. Boca Raton, FL: CRC Press.

Ochi, M., Kashiwazaki, S., Hiratsuka, K., Namba, S. & Tsuchizaki, T. (1992). Nucleotide sequence of the 3´-terminal region of potato virus T RNA. Annals of the Phytopathological Society of Japan 58, 416–425.

Ohshima, K., Nakaya, T., Matsumura, T., Shikata, E. & Kimura, I. (1993). Nucleotide sequences of coat protein and 17K protein genes for a potato leafroll virus Japanese isolate. Annals of the Phytopathological Society of Japan 59, 204–208.

Revill, P. A., Davidson, A. D. & Wright, P. J. (1994). The nucleotide sequence and genome organization of mushroom bacilliform virus: a single-stranded RNA virus of Agaricus bisporus (Lange) Imbach. Virology 202, 904–911.

Roux, L. (1994). Defective-interfering viruses. In Encyclopedia of Virology, vol. 1, pp. 320–323. Edited by R. G. Webster & A. Granoff. New York: Academic Press.

Rozanov, M. N., Koonin, E. V. & Gorbalenya, A. E. (1992). Conservation of the putative methyltransferase domain: a hallmark of the 'Sindbis-like' supergroup of positive-strand RNA viruses. Journal of General Virology 73, 2129–2134.

Saldarelli, P., Minafra, A. & Martelli, G. P. (1996). The nucleotide sequence and genomic organization of grapevine virus B. Journal of General Virology 77, 2645–2652.

Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989). Molecular Cloning: A Laboratory Manual, 2nd edn. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory.

Sato, K., Yoshikawa, N. & Takahashi, T. (1993). Complete nucleotide sequence of the genome of an apple isolate of apple chlorotic leaf spot virus. Journal of General Virology 74, 1927–1931.

Shapira, R., Choi, G. H. & Nuss, D. L. (1991). Virus-like genetic organization and expression strategy for a double-stranded RNA genetic element associated with biological control of chestnut blight. EMBO Journal 10, 731–739.

Shirako, Y. & Wilson, T. M. (1993). Complete nucleotide sequence and organization of the bipartite RNA genome of soil-borne wheat mosaic virus. Virology 195, 16–32.

Solovyev, A. G., Novikov, V. K., Merits, A., Savenkov, E. I., Zelenina, D. A., Tyulkina, L. G. & Morozov, S. Yu. (1994). Genome characterization and taxonomy of Plantago asiatica mosaic potexvirus. Journal of General Virology 75, 259–267.

Song, S. I., Song, J. T., Kim, C. H., Lee, J. S. & Choi, Y. D. (1998). Molecular characterization of the garlic virus X genome. Journal of General Virology 79, 155–159.

Sumi, S., Tsuneyoshi, T. & Furutani, H. (1993). Novel rod-shaped viruses isolated from garlic, Allium sativum, possessing a unique genome organization. Journal of General Virology 74, 1879–1885.

Thompson, J. D., Higgins, D. G. & Gibson, T. J. (1994). CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Research 22, 4673–4680.

Vilches, S. & Castillo, A. (1997). A double-stranded RNA mycovirus in Botrytis cinerea. FEMS Microbiology Letters 155, 125–130.

Vogel, H. J. (1964). Distribution of lysine pathways among fungi: evolutionary implications. American Naturalist 98, 435–446.

Weiller, G. F. & Gibbs, A. (1995). DIPLOMO: the tool for a new type of evolutionary analysis. Computer Applications in the Biosciences 11, 535–540.

White, K. A., Bancroft, J. B. & Mackie, G. A. (1992). Coding capacity determines in vivo accumulation of a defective RNA of clover yellow mosaic virus. Journal of Virology 66, 3069–3076.

Wu, S. X., Rinehart, C. A. & Kaesberg, P. (1987). Sequence and organization of southern bean mosaic virus genomic RNA. Virology 161, 73–80.

Yokoi, T., Takemoto, Y., Suzuki, M., Yamashita, S. & Hibi, T. (1999). The nucleotide sequence and genome organization of Sclerophthora macrospora virus B. Virology 264, 344–349.

Yoshikawa, N., Sasaki, E., Kato, M. & Takahashi, T. (1992). The nucleotide sequence of apple stem grooving capillovirus genome. Virology 191, 98–105.

Yoshikawa, N., Iida, H., Goto, S., Magome, H., Takahashi, T. & Terai, Y. (1997). Grapevine berry inner necrosis, a new trichovirus: comparative studies with several known trichoviruses. Archives of Virology 142, 1351–1363.

Zavriev, S. K., Kanyuka, K. V. & Levay, K. E. (1991). The genome organization of potato virus M RNA. Journal of General Virology 72, 9–14.

JGV Direct table of contents

© 2000 SGM

This article is now available in the January 2001 print issue of JGV (vol. 82, 67–78). The complete issue of the journal may be seen in electronic form on JGV Online.