Virus Genes 29:2, 279–285, 2004 2004 Kluwer Academic Publishers. Manufactured in The Netherlands.
Extensive Variation of Sequence within Isolates of Grapevine Virus B+ BU-JUN SHI,1 NUREDIN HABILI,* RONI GAFNY2 & ROBERT H. SYMONS Waite Diagnostics, School of Agriculture and Wine, Waite Campus, University of Adelaide, Glen Osmond, SA 5064, Australia; 1 Australian Centre for Plant Functional Genomics, Waite Campus, University of Adelaide, Glen Osmond, SA 5064, Australia 2 Institute of Plant Protection, Agricultural Research Organization, The Volcani Center, P.O.Box 6, Bet Dagan 50250, Israel Accepted May 13, 2004
Abstract. Four regions covering 1247 nucleotides of the RNA genome of 20 isolates of a Vitivirus, Grapevine virus B (GVB), from three countries were analyzed. All the regions in these isolates varied in sequence as compared to the published GVB sequence. Of these, the intergenic region varied the most, with 73.2% nucleotide sequence homology, while ORF4 encoding coat protein varied the least when compared both at nucleotide sequence (80.3% homology) and at amino acid sequence levels (90.6% homology). The variations were scattered along each region length and were higher at the nucleotide level than at the amino acid level, but none resulted in a frame shift or stop codon. These results indicate that GVB may exist as a heterogeneous population, possibly resulting from mixing different strains by grafting practices or by RNA-RNA recombination in the grapevine, the only known natural host of this virus. Although it has been reported that GVB is associated with corky bark disease, no corky bark symptoms were observed in any of the GVB positive grapevine sample collected from Australia. Key words: corky bark disease, Grapevine virus B, sequence variation
Introduction Vitivirus is a genus of plant RNA viruses [1], which contains four definite members, GVB, Grapevine virus A (GVA), Heracleum latent virus and Grapevine virus D (GVD). Grapevine virus C (GVC) is a putative member of this genus. These viruses are restricted to the phloem tissue of infected grapevines and two of them (GVA and GVB) are transmitted by mealybugs [2]. However, the main route of transmission of these viruses is by propagation using infected plant material. Vitiviruses possess a single-stranded *Author for all correspondence: E-mail:
[email protected] + The nucleotide sequences reported in this paper have been deposited in EMBL as Accession numbers AY490134AY490174.
positive sense RNA and encode five genes (Fig. 1). GVB was first described in 1993 [2], but corky bark, a serious disease, apparently caused by GVB, was described in 1954 [3], In 1996, the genome of an Italian isolate of GVB was completely sequenced [4]. GVB was first detected in Australia in 1999 in cane samples of symptomless Jade Seedless table grapes [5]. Since then, this virus has been detected in a number of grapevine varieties and rootstocks growing in Australia. Due to its importance for the viticulture industry and the need for a sensitive molecular assay for the detection of all isolates of the virus, we made a comparative study of selected regions of the genomic RNA of a range of GVB isolates. Here, we report results of sequence variation within targeted regions in 20 virus isolates occurring in three countries, Australia, Italy and Israel.
280
Shi et al.
Fig. 1. Genome organization of GVB showing the four sequenced regions as shaded boxes. ‘‘O’’ at the 5¢-end represents the cap structure while ‘‘(A)n’’ at the 3¢-end represents the poly (A) tail, The length of the genome is 7599 nt excluding the poly (A) tail.
Materials and Methods Virus Isolates A total of 20 GVB isolates from different grapevine varieties or from individual vines of the same variety were selected as reference vines (Table 1) and sampled for RNA extraction and sequence analysis. Four of them were from Italy designated ltal1–4. Ital1 was grown in Nicotiana occidentalis in Italy and sent as purified RNA by A. Minafra. This isolate was used as our standard for GVB, because it is the only isolate that
has been fully sequenced [4]. Three isolates were from three vines of Thompson Seedless variety sent from Israel (Golan Heights) and designated as Isr1–3. All the Thompson Seedless vines showed corky bark symptoms. Isolates Is2 and Is3 also tested positive for GVA (unpublished results). Thirteen isolates were from different vineyards in South Eastern Australia designated Aus1–13. Four of these isolates were different varieties of Vitis rootstocks collected from Barossa Valley, South Australia, and 9 were samples of Vitis vinifera. Aus 1–3 were from the Jade Seedless variety.
Table 1. Sources of the GVB isolates and their nucleotide (nt) and aminoacid (aa) identifies (%) in the intergenic region (IR) and open reading frame (ORF)5 Nucleotide identity
Amino acid identity
Isolate
Variety/plant
Location
IR (41 nt)
ORF5 (243 nt)
ORF5 (80 aa)
Aus 1 Aus 2 Aus 3 Aus 4 Aus 5 Aus 6 Aus 7 Aus 8 Aus 9 Aus 10 Aus 11 Aus 12 Aus 13 Ital 1 Ital 2 Ital 3 Ital 4 Isr 1 Isr 2 Isr 3
Jade Seedless-Q Jade Seedless-V Jade Seedless-S Flame Seedless unknowna cv.1 unknowna cv.2 Chardonnay Semillon Nebbiolo Teleki 5C rootstock 110 Richter rootstock 5BB Kober rootstock 1103 Paulsen rootstock Nicotiana occidentalis unknowna cv.1 unknowna cv.2 420A rootstock Thompson Seedless Thompson Seedless Thompson Seedless
Queensland Victoria South Australia Victoria Quarantine Quarantine Victoria South Australia New South Wales South Australia South Australia South Australia South Australia Italy, Standard isolate Bari, Italy Bari, Italy Bari, Italy Vine 1, Israelb Vine 2, Israelb Vine 3, Israelb
75.6 78 78.6 76.7 75.6 73.2 75.6 78 73.2 100 100 100 97.6 100 75.6 78.6 75.6 97.6 92.7 92.7
80.6 80.6 81 80.6 84.3 81 89.7 76.4 76 98.8 99.2 99.2 96.2 100 80.6 87.1 77.7 96.3 93.7 96.3
83.8 83.8 85 85 90 87.5 80 87.5 75 97.5 100 100 94.3 100 86.3 86.4 84.9 100 89.9 90
a
Senders did not reveal the name of these grapevine varieties. These three Thompson Seedless samples were showing corky bark symptoms. In addition, samples 2 and 3 tested positive for GVA by ELISA (R. Gafny, unpublished). b
281
Genome Variation of Grapevine Virus B Table 2. Primer sequences used for RT-PCR to amplify regions of the GVB genome Primer Sequences (5¢-3¢)
Primer positiona
Primer usage
ACACTGATCTTGACAAATGGTGTA TCGGATCCTTCACAATGCCATb TGACCTTCGTAACTGATGCT GCTGTGAAGACGTTCTTAGCAC CTATATCTCAACAGACTGCTCA TCACCTAAATACTTAGACAT ATGTCTAAGTATTTAGGTGA TCAGGCTTCCCAATTCT
nt nt nt nt nt nt nt nt
Amplify Amplify Amplify Amplify Amplify Amplify Amplify Amplify
a
4397–4420 4867–4847 6509–6528 7007–7886 7038–7059 7098–7079 7079–7098 7322—7306
ORF1 ORF1 ORF4 ORF4 IR IR ORF5 ORF5
nt Positions as in Genbank Accession No.X75448. The primers complementary to the genome sequence are indicated by italics.
b
Total RNA Extraction and RT-PCR Total RNA was extracted using the Qiagen RNeasy Plant Mini kit (Qiagen, Germany) as described [6]. RT-PCR was performed according to MacKenzie et al. [7]. Briefly, 1 ll (approximately 1 lg) of total RNA was added to 9 ll aliquots of an RT-PCR reaction mix containing 0.4 lM of each primer (Table 2), 0.4 units Taq DNA polymerase (Invitrogen, USA), 7.2 units Superscript II reverse transcriptase (Invitrogen, USA), 1.8 mM MgCl2, 0.2 lM dNTP, 67 mM Tris (pH 8.9), 50 mM KCl, 6 lM, 0.1% Triton X-100 and 5 mM DTT. The RT-PCR program was as follows: an initial incubation at 50C for 45 min to allow for the reverse transcription, then one minute at 94C. This step was followed by 35 cycles of 94C for 30 s, 56C for 45 s and 72C for 60 s, and finally, an extension time of 5 min at 72C. The PCR products were analysed in a 1.5% agarose gel containing 0.5 lg/ml ethidium bromide and visualised under UV light. Cloning and Sequencing The PCR products were excised from the gel and purified using the Qiagen Gel Purification kit (Qiagen, Germany). The purified RT-PCR products were then cloned into the pGEM-T Easy Vector (Promega, USA), DNA sequencing was carried out by deoxynucleotide chain termination primed with SP6 and T7 primers from the vector. A total of 5 clones per sample were sequenced. The nucleotide sequences were analysed using the
Geneworks (IntelliGenetics, Mt View, CA, USA) and BLAST programs [8].
Results Sequence Variation in all the Four Regions of all the Isolates Four regions on the GVB genome of each isolate were targeted using GVB specific primers (Table 2) designed from the published sequence with the GenBank Accession number: X75448 [4]. Here, this isolate has been named the GVB standard isolate. The selected targets included parts of the open reading frame (ORF) 1 encoding RNA polymerase (RdRpol), the ORF4 encoding coat protein, the ORF5 encoding RNA-binding protein and a unique intergenic region (IR) between ORFs 4 and 5, which may act as a promoter for the ORF5 (see the targeted regions highlighted in Fig. 1). A total of 1247 nucleotides (nt) covering 16.4% of the entire genome across the four regions were sequenced. The sequence comparison analysis showed that none of the isolates were 100% identical to each other (Table 1). When we sequenced an RNA sample from the standard isolate of GVB [4] sent from Italy (Ital1), we observed that there was one nucleotide substitution in ORF1 (data not shown). We questioned whether the single nucleotide substitution resulted from a Taq polymerase-induced error [9], or a reverse transcriptase-induced error [10]. The result of a repeated RT-PCR and cloning experiment carried out on the same region (ORF1) confirmed the
282
Shi et al.
Table 3. Nucleotide (nt) and amino acid (aa) identifies (%) of ORFs 1 and 4 of the 20 GVB isolates Nucleotide identity
Amino acid identity
ORF1 Isolatea Variety/plant (471 nt)
ORF4 ORF1 ORF4 (492 nt) (156 aa) (163 aa)
Aus2 Aus5 Aus8 Ital1
81.1 80.3 81.4 99.8
99.6 81.3 83.6 100
95.5 95.5 96.8 99.4
100 97.6 90.6 100
98.1
98
99.4
98
93.1
99.5
88.5
100
96.2
98.8
88.5
Isr1 Isr2 Isr3 a
Jade Seedless unknown Semillon Nicotiana occidentalis Thompson Seedless Thompson Seedless Thompson Seedless
97.5
Only these isolates gave specific RT-PCR products with primers specific to ORFs 1 and 4 of GVB, which were sequenced. The other isolates except Aus9 gave the specific RT-PCR products, but were not sequenced. The isolate Aus9 gave no RT-PCR product with the primers specific to ORFs 1 and 4 of GVB.
single nucleotide difference. It is possible that this RNA sample was from a different batch (passage) of the virus in N.occidentalis. The variations at both nucleotide and amino acid levels in each region among all the 20 isolates were calculated by comparing to the published sequenced isolate (Tables 1 and Fig. 2). Overall the variations of amino acid sequence for all the four regions of most isolates were much less than those of nucleotide sequence. Of the four regions shown in Tables 1 and 3, the IR varied the most at the nucleotide level, while ORF4, which encodes the coat protein gene, varied the least both at nucleotide and amino acid sequence levels. This suggests that the IR may be under less selective pressure as compared to the CP gene. This is in contrast to the recent finding by Goszczynki and Jooste [11] who observed higher sequence variability in the CP gene of GVA, the type member of vitiviruses. The Isolates Fall into Two Groups Based on Sequence Homology Based on the degree of sequence variations shown in Tables 1 and 3, the 20 isolates fall into two groups: those with a sequence close (90% and over) to that of the published (standard) GVB and those
relatively distant from this isolate. Group I includes four isolates from different (Fig. 3) Vitis rootstock varieties from South Australia and the three isolates from Israel (Table 1, Fig. 3). The isolates in this group shared over 90% sequence homology at the nucleotide level with the published sequence, suggesting a common origin possibly belonging to the same strain. Group II includes the remaining 13 isolates, which shared a sequence homology of less than 90% with the published sequence. Some of these isolates varied significantly. One isolate, Aus9, had the most divergent sequence both at nucleotide and amino acid levels in all the four regions compared to the published GVB sequence, (see Tables 1 and 3). For example, in the IR region this isolate only shared 73.2% nucleotide sequence homology with the published GVB. In the OFR5 region, its homology was 76% at the nucleotide level and 75% at the amino acid level. It is interesting to note that Aus9 failed to give any RT-PCR product with the primers specific for either ORF1 or ORF4 (data not shown). These data indicate that the Aus9 isolate may be a divergent strain of GVB. For a further definition of this isolate, full sequence information and other biological data are needed. Generally, the isolates of group II may represent different strains of GVB in term of having less sequence homology with the standard virus. It is possible that these strains have resulted from RNA-RNA recombination between GVB and other related viruses such as GVA and GVD following a co-infection event. GVB, GVA and GVD can, co-exist in grapevines (Table 1, isolates Is2 and 3 and N. Habili, unpublished). The proposed possible recombination between GVB and other related viruses is supported from the sequence of the RdRpol region in which a stretch of 46 nt of Aus6 isolate matched with that of GVA sequence [12, GenBank Accession No. AY244516] and likewise a stretch of 31 nt of Aus7 isolate matched with that of GVD (unpublished results). The assignment of the 20 isolates into two groups is consistent with the phylogenetic analysis based on the ORF5 nucleotide sequences (Fig. 3). Taken together, our results suggest that GVB is a virus with high sequence variability and different strains may exist in this virus.
Genome Variation of Grapevine Virus B
283
Fig. 2. Positions of the varied nucleotide and amino acids residues in each of the four sequenced regions of the GVB genome.
Despite the presence of significant variations, especially in ORF5 of all the isolates, an 8-mer in this region, ATGTCTAA, was conserved in all these isolates. This sequence core in the ORF5 could be used to design primers for the possible detection of GVB isolates. A similar size primer is present in the NIb gene of all known potyvirids, which can be useful for their identification [13]. Nucleotide Sequence Variation did not Change Any Encoded Protein The variations observed occurred at both nucleotide and amino acid levels at positions covering the entire length of each region (Fig. 3). In some positions, nucleotide substitution, or the mutation frequency was very high. For example, 15 isolates out of 20 had
a varied nucleotide at position 148 of ORF 5, indicating that this position may be under selection pressure. However, interestingly, the variations observed in these 20 isolates did not include any deletion or insertion of a nucleotide and thus did not result in any frame shift. In addition, none of the mutations observed in this study resulted in a stop codon or coded for a new protein, suggesting importance of the existing proteins in the virus biology. GVB is Widely Present in Symptomless Australian Grapevines Using primers designed from ORF5 (Table 2), we have developed a nested PCR assays for the detection of GVB [5]. Of 4384 grapevine samples sent
284
Shi et al.
Fig. 3. Phylogenetic analysis of nucleotide sequence identities on the ORF5 region from the 20 GVB isolates. The group I isolates that show more than 90% sequence homology with the published GVB (boxed) are in bold. The sequence alignment and branching order were carried out using the GeneWorks programs.
from major viticultural regions of Australia, 232 (5.2%) tested positive using the assay method described above. None of the grapevine plants showed the symptoms of corky bark, a GVB associated disease [2,14]. No symptoms have been observed on LN33 (Vitis berlandieri x Couderc 1,613), a biological indicator for the corky bark disease, following graft inoculation using Australian grapevine varieties [14]. In fact, corky bark is a quarantinable disease in Australia, and any suspected grapevine material brought into the country is destroyed at post-entry quarantine glasshouses [14]. It is possible that corky bark is a complex disease caused by a mixture of viruses. Discussion We analysed four regions on the genome of 20 GVB isolates from three countries and found extensive variation in the sequence of their genome. These isolates could be divided into two groups based on
homology with the sequenced GVB strain from Italy. Group I consists of isolates which shared more than 90% homology with the standard GVB. It is interesting to note that known corky bark associated isolates of Ital 1 and the three Thompson Seedless isolates from Israel (unpublished) are clustered in this group (Fig. 3). Nearly two-third of the isolates had less than 90% homology with the standard GVB, These included most isolates from Australia which are not associated with corky bark [2,14]. A possible explanation for the presence of extensive sequence variation in the GVB RNA may be due to the extensive vegetative propagation of the grapevine, its only natural host, over millennia. This, together with grafting practices has resulted in the mixing of GVB isolates from various sources. This not only may evoke RNA recombination but also may force the virus to induce changes to its sequence in order to be able to adapt to the new environment [15]. The detection of an 8-mer in ORF5 of all the isolate may be useful for designing a universal tool for the detection of divergent strains of GVB (see also: 13). Acknowledgments This project was supported by an Australian ARC-SPIRT grant and by the Australian Grape and Wine Research and Development Corporation. We are grateful to J. W. Randles for his remarks. We wish to thank A. Minafra (Bari, Italy) for supplying us with the GVB isolates, and D. Webb for assisting us with sequence analysis. References 1. Martelli G., Saldarelli P., and Minafra A., 12th Meeting of the International Council for the Study of Viruses and Virus-like Diseases of the Grapevine, 1997, p. 23. 2. Boscia D., Savino V., Minafra A., Namba S., Elicio V., Castellano M.A., Gonsalves D., and Martelli G.P., Archives of Virology 130, 109–120, 1993. 3. Hewitt W.B., California Dept. Agr. Bul. 43, 47–64, 1954. 4. Saldarelli P., Minafra A., and Martelli G.P., J Gen Virol 77, 2645–2652, 1996. 5. Habili N., and Symons R.H., Australian Grapegrower and Winemaker 429, 58–59 1999. 6. Shi B.J., Habili N., and Symons R.H., Ann Appl Biol 142, 349–355, 2003. 7. MacKenzie D.J., McLean M.A., Mukerji S., and Green M., Plant Disease 81, 222–226, 1997.
Genome Variation of Grapevine Virus B 8. Altschul S.F., Gish W., Miller W., Myers E.W., and Lipman D.J., J Mol Biol 215, 403–410, 1990. 9. Braho M.A., Moya A., and Barrio E., J Gen Virol 79, 2921–2928, 1998. 10. Coffin J.M., Genetic Diversity of RNA Viruses. SpringerVerlag Berlin, pp. l43–164, 1992. 11. Goszczynki D.E., and Jooste A.E.C., Eur. J Plant Pathol 109, 397–403, 2003.
285
12. Galiakparov N., Goszczynski D.E., Che X., Batuman O., Bar-Joseph M., and Mawassi M.,Virology 312, 434–448, 2003. 13. Gibbs A.J., Mackenzie A.M., and Gibbs M.J., J Virol Methods 112, 41–44, 2003. 14. Whattam M., Australasian Plant Path 30, 379–380, 2001. 15. Garcia-Arenal F., Fraile A., and Malpica J.M., Ann. Rev Phytopathol 39, 157–186, 2001.
