Single-strand conformation polymorphism (SSCP), cloning and sequencing reveal a close association between related molecular variants of Grapevine virus A (GVA) and Shiraz disease in South Africa

Authors


*E-mail: GoszczynskiD@arc.agric.za

Abstract

When a recently discovered 234-nt fragment of the Grapevine virus A (GVA) genome, located in ORF3, was used in single-strand conformation polymorphism (SSCP), it produced patterns of DNA bands which were indicative of GVA variants of molecular group II. This technique of rapid identification of variants of group II was applied to the study of GVA variants in various grapevines with different Shiraz disease (SD) status. The results, supported by nucleotide sequence data, revealed that GVA variants of molecular group II are closely associated with SD in South Africa, and showed that variants of group III are commonly present in SD-susceptible grapevines that consistently do not exhibit symptoms of this disease.

Introduction

Shiraz disease (SD) (Corbett & Wiid, 1985) is a destructive disease of own-rooted as well as grafted grapevine cvs Shiraz, Merlot, Malbec, Gamay and Viognier in South Africa. Canes of grapevines affected by SD never mature fully. They remain green for a long time during the growing season and exhibit excessive cambium and phloem development, which causes severe longitudinal cracks of non-lignified bark. The wood of SD-affected canes is weakly developed, making the plants rubbery in texture. The disease affects growth, delays or totally hampers budburst and severely affects the production of fruit. Once a grapevine shows definite SD symptoms, it never recovers and dies within 3–5 years. Other cultivars, although infected (SD-positive), do not exhibit symptoms. The disease is easily transmitted from them to disease-free plants by grafting of infected tissue.

Shiraz disease was successfully transmitted between grapevines using mealybugs (Planococcus ficus) and it was found that SD-affected plants are always infected with a Vitivirus, Grapevine virus A (GVA) (Goszczynski & Jooste, 2003a). When 2-month-old virus-free cv. Merlot grapevines were subjected to massive infestation with mealybugs (more than 100 insects per plant), which had earlier fed on plants affected by Shiraz disease, symptoms of the disease were immediately visible in the first newly growing shoots (Goszczynski & Jooste, 2003a). The same was observed after grafting of SD-positive grapevines to SD-susceptible plants of Shiraz and Merlot. However, it is not known how long the disease takes to develop when inoculum is limited. A similar disease, called Australian Shiraz disease (AuSD), occurs in Australian vineyards, where it has been reported that this disease is also associated with GVA (Habili & Randles, 2004). The virus has positive-sense single-stranded RNA of 7349–7351 nucleotides, excluding a poly(A) tail at the 3′ terminus (Minafra et al., 1997; Martelli et al., 1997; Galiakparov et al., 2003a). Its genome is organized into five open reading frames (ORF1-5) (Minafra et al., 1997). ORF1, 3, 4 and 5 encode putative replicase, movement protein (MP), capsid protein (CP) and nucleic-acid-binding proteins, respectively (Minafra et al., 1997; Galiakparov et al., 2003b). The function of a protein encoded by ORF2 is not known. GVA can be transmitted by mechanical inoculation and by the insect vector, P. ficus, to the herbaceous host, Nicotiana benthamiana, which markedly facilitates the study of the virus. The virus is extensively molecularly heterogenic. Divergent variants of the virus, which cluster into three molecular groups (I, II and III) based on nucleotide similarity in the 941- to 942-nt 3′ terminal part of the virus genome, were identified (Goszczynski & Jooste, 2003b). Variants share 91·0–99·8% nt similarity within groups and 78·0–89·3% nt similarity between groups in that genomic region. Results suggested that mixed infections of divergent GVA variants are common in GVA-infected grapevines in vineyards in South Africa (Goszczynski & Jooste, 2003c). Sequence analysis of the 3′ terminal region of GVA variants in SD-positive and SD-affected grapevines (6–10 clones per infected plant) did not associate a specific group of variants with the disease. Although, according to this limited-scale study, variants of group III were dominant (only single sequences of other groups were sometimes detected) in all three sources of SD-positive grapevines used as controls in woody indexing (transmission of the disease to SD-susceptible grapevines by grafting) by the SA grapevine industry, the variants of group II were clearly dominant in two SD-affected field-collected plants of Merlot (Goszczynski & Jooste, 2003a). GVA was also detected in many Shiraz plants, which consistently did not exhibit SD symptoms (Goszczynski & Jooste, 2003a). Sequencing of the 3′ terminal region of 10 clones of GVA from one of these SD-negative plants, named Shiraz GTR1, revealed only a single variant of molecular group III. Later, a variant of group II was mechanically transmitted from this grapevine to N. benthamiana (Goszczynski & Jooste, 2002). One of the major limitations in these studies is the need to rely on cloning and sequencing alone to identify molecular variants of GVA in grapevines. While a method based on RT-PCR was developed for rapid detection of the most divergent GVA variants of group III (Goszczynski & Jooste, 2003b), there was no technique available that could differentiate between variants of groups I and II. Single-strand conformation polymorphism (SSCP) (Orita et al., 1989) was previously used for rapid analysis of molecular heterogeneity of GVA (Goszczynski & Jooste, 2002). It was able detect small nucleotide differences among variants of the virus, but it did not allow the estimation of how extensive these differences were, or to which molecular group a particular variant belonged. More recently, a fragment of ORF3 of GVA was identified which in SSCP analysis revealed patterns of DNA bands indicative of GVA variants of molecular group II. The application of the SSCP technique to the study of the association of GVA variants with Shiraz disease, supported by cloning and sequencing, is presented in this paper.

Materials and methods

GVA-infected, SD-negative and SD-affected cv. Shiraz and SD-affected cv. Merlot grapevines and one SD-affected cv. Viognier grapevine were collected from various vineyards in the Western Cape, South Africa. SD-positive cv. Cinsaut Blanc clone P163/12 was obtained from the Directorate of Plant Health and Quality, Stellenbosch. The terms ‘SD-affected’ and ‘SD-positive’ used in this paper refer to grapevines of SD-susceptible cultivars that showed symptoms of the disease, and grapevines infected with SD but not susceptible to the disease, respectively. In SD-positive grapevines the disease is latent and can be easily transmitted to disease-free plants by grafting of infected tissue. Mechanical and mealybug (P. ficus) transmission of GVA from grapevines to N. benthamiana, isolation of dsRNA, RT-PCR, SSCP, purification of DNA products of RT-PCR from agarose, cloning, sequencing and sequence analysis were carried out as described by Goszczynski & Jooste (2003b,c). In RT-PCR, GVA-specific primer pairs MP/CPdt (5′-TGCCAGAGGTGTTTGAGACAAT-3′/5′-TTTTTGTCTTCGTGTGACAACCT-3′) (De Meyer et al., 2000), H7038/C7273 (5′-AGGTCCACGTTTGCTAAG-3′/5′-CATCGTCTGAGGTTTCTACTA-3′) (MacKenzie, 1997) and MP2F/MP1R (5′-TCTGAACAAGGCCCTGCA-3′/5′-AGATTCTTGCCATGGGGCAT-3′) (this work) were used for amplification of the 3′ terminal region of the virus (PCR product 986 bp, complementary to part of ORF3, ORF4, ORF5 and part of 3′NTR), ORF5 (PCR product 236 bp) and ORF3 (PCR product 234 bp), respectively. For SSCP of GVA from grapevines, the virus was RT-PCR-amplified from purified dsRNA using primer pair MP2F/MP1R. A 1-µL aliquot of the 234-bp DNA product was mixed with 9 µL Bromophenol Blue loading solution (Promega), incubated at 99°C for 10 min, cooled on ice for 2 min and electrophoresed in 15% acrylamide/bis-acrylamide (29·2/0·8; w/w), 0·75-mm gels in 0·5× TBE buffer at 5°C for 2 h at 200 V, using a Mini-protean II dual slab cell (Bio-Rad). For SSCP of GVA from N. benthamiana, 234-bp products of RT-PCR amplification were purified from low-melting-point agarose using the Wizard PCR Preps DNA Purification System (Promega). For SSCP of cloned 234-bp fragments of GVA, randomly selected bacterial colonies were plated on Luria-Bertani medium with 100 µg ampicillin mL−1, and after overnight incubation at 37°C a small amount of new bacterial growth was re-suspended in 200 µL sterile distilled water by vortexing, and then 1 µL bacterial suspension was used in PCR. One µL of PCR product mixed with 9 µL loading buffer was analysed by SSCP. Phylogenetic analyses of nucleotide sequences were carried out using the dnaman version 5·2·9 (Lynnon Biosoft) software package. Sequences of primers were excluded from the analyses. Phylogenetic trees were constructed with the neighbour-joining method (Saitou & Nei, 1987) using evolutionary distances calculated according to the method of Kimura (1980). Bootstrap analysis of the data, based on 1000 permutations, was used to assess the statistical confidence of the topologies. Only bootstrap values higher than 50% are shown. The analyses include the sequence data of GVA isolate Is151 (Minafra et al., 1997) and the sequence of Grapevine virus B (GVB) (Sardarelli et al., 1996) deposited in the GenBank/EMBL database with the accession numbers X75433 and X75448, respectively. In every phylogenetic analysis sequence data for GVB were included as an outgroup.

Results and discussion

Differentiation of GVA variants of molecular group II using SSCP analysis of a 234-nt fragment of ORF3

Recent comparative molecular analysis of the movement protein gene (ORF3) of single GVA variants of the different molecular groups transmitted from various grapevines to N. benthamiana, shown in Table 1 and Fig. 1 (DEG, unpublished data), revealed a 234-nt sequence that in SSCP analysis appeared to differentiate variants of molecular group II from those of groups I and III. The sequence was located in the 5′ terminal half of ORF3 and was easily RT-PCR-amplified using primer pair MP2F/MP1R. In general, the SSCP bands of the 234-nt sequence of variants of molecular group II migrated slowly in polyacrylamide gels and were diffuse in appearance (Fig. 2). The reason for the diffuse appearance of bands was not clear, but these results were reproducible under the conditions used in this study. Although further study of the variants from various grapevines revealed that just one nucleotide substitution could drastically change the appearance of the bands, an example of which is shown in Fig. 3, it was a very rare occurrence (see further results). The potential for this technique to be used for rapid identification of variants of molecular group II in grapevines was further investigated.

Table 1. Grapevine virus A (GVA) variants from three molecular groups isolated by transmission from various grapevines with differing Shiraz disease (SD) status to Nicotiana benthamiana
GVA variantaSource grapevineMethod of transmissionMolecular groupb
CultivarSD-status
  • a

    SSCP analysis of the 3′ terminal part of variants, RT-PCR-amplified using MP/CPdt primers (see Materials and methods) and digested with restriction enzyme DdeI, supported by sequence data of 3–5 clones of this genomic region per infected N. benthamiana plant, strongly suggested that these are single variants of the virus (Goszczynski & Jooste, 2002; Goszczynski, unpublished results).

  • a,b

    Molecular groups of GVA were determined on the basis of phylogenetic analysis of the 3′ terminal 941- to 942-nt sequences of variants of the virus (see Fig. 1). All sequence data for the isolates shown in this table, including the 234-nt fragment of ORF3 and 941-to 942-nt 3′ terminal part of the virus used in this work, were deposited in the GenBank/EMBL database with accession numbers: (1) AF441234; (2) DQ855084; (3) DQ855085; (4) AF441235; (5) DQ855086; (6) DQ855081; (7) DQ855087; (8) DQ855083; (9) DQ855082; (10) DQ855088; (11) DQ787959.

1. 92·778C. SauvignonSD-positivemechanicallyI
2. GTG11-1ShirazSD-negativemechanicallyI
3. MSH18-1ShirazSD-affectedmechanicallyI
4. JP98-1Waltham CrossSD-positivemechanicallyII
5. GTR1-2ShirazSD-negativemechanicallyII
6. GTR1SD-1ShirazSD-affectedmechanicallyII
7. BMo32-1MerlotSD-affectedmechanicallyII
8. KWVMo4-1MerlotSD-affectedmechanicallyII
9. P163-M5C. BlancSD-positivemealybugsII
10. P163-1C. BlancSD-positivemechanicallyIII
11. GTR1-1ShirazSD-negativemechanicallyIII
Figure 1.

Phylogenetic tree depicting three molecular groups (I, II, III) of Grapevine virus A (GVA) variants, based on the alignment of 941- to 942-nt 3′ terminal parts of the virus genome. The sequence of GVA isolate Is151 from Italy (underlined) was included in the analysis. The tree was constructed using dnaman version 5·2·9 (Lynnon Biosoft).

Figure 2.

SSCP analysis of 234-nt fragments of movement protein of various Grapevine virus A (GVA) variants of three molecular groups (I, II, III) isolated in Nicotiana benthamiana. Variants are described in Table 1.

Figure 3.

Examples of 234-nt sequences of ORF3 of the Grapevine virus A (GVA) variant of molecular group II, differing by one nucleotide (a), and revealing different SSCP patterns (b).

Molecular variants of GVA in field-collected grapevines with different SD status

The virus was amplified by RT-PCR, using primers MP2F/MP1R, from 14 SD-negative and 24 SD-affected cv. Shiraz grapevines, which were collected from the same vineyard. Other SD-affected grapevines used in the study were from two vineyards of cv. Shiraz (three plants per vineyard), five plants of cv. Merlot from a single vineyard and one plant of cv. Viognier (Fig. 4d). SSCP analysis of the 234-bp RT-PCR products revealed slowly migrating bands that were diffuse in appearance, indicative of GVA variants of molecular group II associated with the majority of grapevines affected by Shiraz disease (Fig. 4b–d), but sporadic in SD-negative vines (Fig. 4a). Multiple SSCP bands, which were observed in GVA from SD-negative as well as SD-affected grapevines, indicated that individual grapevines were infected with more than one molecular variant of the virus. To identify the variants, 234-bp RT-PCR products from nine selected SD-negative and nine SD-positive grapevines (Fig. 4a,d) were cloned and sequenced. The clones were amplified by PCR directly from bacterial colonies using MP2F/MP1R primers and the products of amplification were analysed by SSCP (Fig. 4e,f). Clones representing different SSCP patterns were sequenced (Fig. 4g). Combined results of SSCP and sequence analyses suggested that in all SD-negative grapevines, GVA variants of group III predominate, except in one grapevine (Fig. 4e,g, samples 3·1 and 3·4) which was infected only with variants of molecular group II. SD-affected grapevines were infected with mixtures of variants of molecular groups II and III (Fig. 4f,g), except for the SD-affected cv. Viognier grapevine, which appeared to be infected only with a variant of group II (Fig. 4f,g, sample 9·1SD). These results suggest that GVA variants of molecular group II are closely associated with Shiraz disease and variants of molecular group III appear to dominate in SD-susceptible grapevines that do not exhibit symptoms of this disease in South Africa.

Figure 4.

SSCP profiles of 234 nt fragments of MP of Grapevine virus A (GVA) from various field-collected Shiraz disease (SD)-negative (a) and SD-affected (b–d) grapevines. DNA fragments of GVA isolates with SSCP profiles marked with numbers 1 to 9 were cloned and then five clones per isolate were analysed using the SSCP technique (e, f). Clones with distinct SSCP patterns, marked with bold numbers, were sequenced. Evolutionary relationships of sequences were determined using phylogenetic analysis (g). The sequences of GVA isolates GVAIs151 (underlined), BMo32-1 and P163-1 (shaded), representing molecular groups I, II and III, respectively, were included in the analysis.

Molecular variants of GVA in grapevine cvs Merlot and Shiraz with differing SD status, propagated from field-collected plants that originally were symptomless or exhibited only mild symptoms

Further evidence that variants of group II are closely associated with Shiraz disease was obtained in the study of SD-negative and SD-affected cv. Merlot grapevines, which were established in the laboratory from cane cuttings from two Merlot plants (BMo32 and KWVMo4) that exhibited very mild SD symptoms, collected from two different vineyards in 2000. As was described earlier by Goszczynski & Jooste (2003a), GVA was easily detected (strong ethidium bromide staining of DNA products of RT-PCR amplification of the virus) only in SD-affected plants. The virus was also present in SD-negative plants, but surprisingly, it was consistently difficult to detect. The plants were tested using RT-PCR, with oligonucleotide primer pair H7038/C7273 complementary to sequences of ORF5 of GVA, every year over a few years. In 2005, the same dsRNA samples, extracted from these grapevines in 2003, were again tested for the presence of GVA using the same RT-PCR technique. Results confirmed those described above. Strongly ethidium bromide-stained DNA bands were observed only for GVA, which was RT-PCR amplified from SD-affected plants (Fig. 5a). The dsRNA samples were also tested using RT-PCR, with primers MP2F/MP1R complementary to ORF3 of GVA. The virus was easily detected in SD-affected and in most SD-negative plants using these primers (Fig. 5b). SSCP analysis of the 234-bp DNA products revealed the apparent presence of variants of molecular group II in SD-affected plants and showed multiple bands for GVA from SD-negative plants (Fig. 5c). Selected RT-PCR products from SD-affected and SD-negative grapevines were cloned, and clones with differing SSCP patterns were sequenced. SSCP analysis of 26 clones of GVA from SD-affected Merlot BMo32 revealed 22 clones with the SSCP pattern typical of a variant of molecular group II (Fig. 6a,e, sample 1) and four clones with the SSCP pattern of a variant of molecular group III (Fig. 6a,e, sample 2). Thirteen clones of GVA from SD-negative Merlot BMo32 revealed patterns of different GVA variants of molecular group III only (Fig. 6b,e). Analogous results were obtained for GVA, which was cloned from SD-affected and SD-negative Merlot KWVMo4 (Fig. 6c,d). SSCP analysis of 56 clones of GVA from SD-affected plants revealed 54 clones with a SSCP pattern typical of a molecular variant II of the virus (Fig. 6c,e, sample 7). The SSCP pattern of only two clones was different (Fig. 6c, sample 8). Sequence data revealed that it was also a variant of molecular group II, which differed only by one nucleotide substitution (C to T at position 105 of the sequence) from the dominant variant with an SSCP pattern typical of molecular group II (Fig. 6e). Of 12 clones of GVA from SD-negative Merlot KWVMo4, 11 revealed patterns of different variants of GVA of molecular group III (Fig. 6d,e, samples 9, 11–15). One cloned sequence was from a GVA variant of molecular group II, showing the typical SSCP pattern for this group of variants (Fig. 5d,e, sample 10), and one sequence of a variant of molecular group I (Fig. 6d,e, sample 16). These results indicate an uneven distribution of GVA variants in infected plants. The GVA variants of molecular groups II and III, which infected both field-grown mother plants of Merlot, were separated into different plants after propagation from cuttings. The presence of only GVA variants of group II in SD-affected Merlot KWVMo4 plants, or clear domination of variants of group II mixed with variants of group III in plants of Merlot BMo32, support the hypothesis that variants of molecular group II are involved in Shiraz disease. Clear domination or the presence of only variants of group III in SD-negative plants strongly suggest that the variants of this group are not associated with the disease.

Figure 5.

(a, b) RT-PCR amplification of Grapevine virus A (GVA) from Shiraz disease (SD)-affected (1, 5) and SD-negative (2–4, 6–10) cv. Merlot grapevines propagated from two field-collected plants BMo32 (1–4) and KWVMo4 (5–10), which originally did not exhibit clear symptoms of the disease. In RT-PCR, oligonucleotide primers H7038/C7273 (a) and MP2F/MP1R (b), complementary to fragments of ORF5 and ORF3 of GVA, respectively, were used. DNA products of amplification of GVA using primers MP2F/MP1R were analysed using SSCP (c). The products, with SSCP profiles shown in lanes 1, 3, 5 and 9, were cloned. Results of SSCP analysis of the clones are shown in Figure 6.

Figure 6.

SSCP profiles detected in the study of clones of RT-PCR-amplified 234-nt fragments of ORF3 of Grapevine virus A (GVA) from Shiraz disease (SD)-affected (a, c) and SD-negative (b, d) cv. Merlot grapevines. The grapevines were propagated from two field-collected plants, BMo32 (A, B) and KWVMo4 (c, d), which originally did not exhibit clear symptoms of the disease (see also Figure 5). Numbers under the SSCP profiles are numbers of specific profiles detected above numbers of clones analysed. All clones, of which SSCP profiles are shown, were sequenced. Evolutionary relationships of obtained sequences were determined using phylogenetic analysis (e). The sequences of selected GVA variants (shaded), shown in Figure 1, representing all three molecular groups (I, II, III) of the virus, were included in the analysis.

The third line of evidence supporting the above conclusions was obtained in the study of a field-collected GVA-infected cv. Shiraz grapevine. The grapevine, which was clearly affected by leafroll disease (LR) and did not exhibit symptoms of Shiraz disease, named Shiraz GTR1, was collected in 2000. Two GVA variants, named GTR1-1 and GTR1-2, of molecular groups III and II, respectively (Table 1, Fig. 1), were mechanically transmitted from this grapevine to N. benthamiana (Goszczynski & Jooste, 2003b). The plant was propagated from cuttings to nine plants in the laboratory. All established plants consistently did not exhibit symptoms of Shiraz disease for 2 years. In 2003, one of them exhibited symptoms of the disease. The other plants remained SD-free. DsRNA was purified from SD-affected and SD-negative plants every year and tested for the presence of GVA. The SSCP-based method described in this paper was applied to the identification of GVA variants in samples of dsRNA purified from one of the SD-negative plants in 2002 and 2005, and from a SD-affected plant in 2003, 2004 and 2005. The virus was RT-PCR-amplified from each sample using primers MP2F/MP1R, the 234-bp products cloned and the clones analysed by SSCP (Fig. 7a–d). Clones with distinct SSCP patterns were sequenced (Fig. 7e). Results revealed that of 90 clones of GVA from SD-negative Shiraz GTR1 grapevine (Fig. 7c), 88 contained sequences of the variant of group III (Fig. 7c,e, sample 1) and only two were sequences of the variant of group II (Fig. 7c,e, sample 2). The sequences were almost identical to sequences cloned from variants GTR1-1 and GTR1-2, respectively (Fig 7e), transmitted earlier from SD-negative Shiraz GTR1 grapevine to N. benthamiana (Goszczynski & Jooste, 2002, 2003b). Of 131 clones of GVA from SD-affected Shiraz GTR1 grapevine (Fig. 7d), 50 clones contained sequences of a variant of molecular group II, showing a SSCP pattern characteristic of GVA variants of this group (Fig. 7d,e, sample 3). The sequence was identical to that of GVA variant GTR1SD-1 (Fig. 7e), which was transmitted from the above grapevine to N. benthamiana (Table 1). Another 68 and eight clones were sequences of a variant of molecular groups III and I, respectively (Fig. 7d,e, samples 4 and 5). Single cases of GVA variants detected in the SD-affected plants were variants of molecular group II, which, although they showed SSCP patterns distinct from the main variant of group II in this grapevine (Fig. 7d, samples 6, 7 and 8), differed only by one nucleotide from this variant (Fig. 3a). As described above, all grapevines were established from cuttings of field-collected GVA-infected grapevine cv. Shiraz, which did not exhibit symptoms of Shiraz disease. The relatively high level of a molecular group II (GTR1SD-1-like) variant in a plant that became SD-affected, and the low level of a group II variant (GTR1-2-like) in a plant that was consistently SD-negative, despite being clearly infected with high levels of a molecular group III variant, strongly suggests that GVA variants of molecular group II are involved in Shiraz disease, and that variants of group III are not pathogenic to SD-susceptible grapevines in South Africa. Transmission of GVA variants of different molecular groups from N. benthamiana to grapevines (fulfillment of Koch's postulates) would help to clarify the involvement of variants in the disease. Unfortunately, although GVA was transmitted from grapevines to N. benthamiana for the first time more than 25 years ago (Conti et al., 1980), all attempts to transmit it back to the grapevine host failed. Successful transmission of GVA back to grapevines of SD-susceptible cultivars would also enable the potential role of other pathogens in the disease to be investigated.

Figure 7.

Typical results of SSCP analysis of clones of RT-PCR-amplified 234-nt fragments of ORF3 of Grapevine virus A (GVA) from Shiraz disease (SD)-negative (a) and SD-affected (b) grapevines cv. Shiraz. The grapevines were propagated from the same field-collected plant, which did not exhibit symptoms of the disease. SSCP profiles, which were detected in the study of 90 clones of GVA from SD-negative and 131 clones of GVA from SD-affected grapevines, are shown in c and d, respectively. Numbers under the SSCP profiles are numbers of specific profiles detected above numbers of clones analysed. All fragments for which SSCP profiles are shown in c and d were sequenced, and evolutionary relationships of sequences was determined using phylogenetic analysis (e). The sequences of selected GVA variants (shaded), shown in Figure 1, representing all three molecular groups (I, II, III) of the virus, were included in the analysis.

Acknowledgement

The author thanks Winetech, South Africa, for partially funding this study.

Ancillary