Grapevine yellows diseases in Spain : eight year survey of disease spread and molecular characterization of phytoplasmas involved by

Among grapevine yellows phytoplasma diseases in Europe, flavescence dorée (FD) is the most devastating and in the last decade has reached Spanish vineyards, mainly in Catalonia. An eight-year survey was carried out in the areas where the disease has spread (Alt Empordà, Catalonia, Northern Spain) and in the remaining vine-growing areas of Catalonia. Sequence analyses of a portion of the 16S-23S ribosomal DNA cistron, from selected grapevine samples from Catalonia, showed that the phytoplasmas involved in grapevine yellows belong to 16S ribosomal subgroups V-D (flavescence dorée, FD) and XII-A (bois noir, BN). A set of Spanish FD isolates collected during these years were further studied by RFLP analyses of the 16S-23S ribosomal DNA fragment, as well as the rpS3 and SecY genes. All the FD phytoplasma strains studied were related to phytoplasmas belonging to ribosomal protein subgroup rp-E.

FD is the most aggressive phytoplasma among those associated with grapevine diseases, and is thus subject to quarantine restrictions.Genetic relatedness of this pathogen, revealed by 16Sr DNA analyses, showed that two different ribosomal subgroups (16SrV-C, also known as FD-C, and 16SrV-D, also known as FD-D) are associated with outbreaks in Italy and both are transmitted by the vine leafhopper insect Scaphoideus titanus Ball (Bertaccini & al., 1997;Martini & al., 1999;Mori & al., 2002).In recent years outbreaks associated with FD-D type phytoplasmas were observed in several regions of Northern Italy and in France (Angelini & al., 2001;Bertaccini, 2002;Boudon-Padieu, 2003).
In Spain S. titanus was identified in the four provinces of Catalonia (NE Spain): Barcelona, Gerona, Lérida and Tarragona (Barrios & al., 1998).The first FD outbreak occurred in Gerona (Laviña & al., 1996), which forced the Catalonian administration to legislate a compulsory program to eradicate FD at Alt Empordà (Rahola & al., 1997).In application of this program, vineyards were sprayed (via helicopter) with a pyrethroid insecticide to control S. titanus in areas ranging from 1900 Ha in 1997 to 2030 Ha in 1999.
In this work, we surveyed ca.63027 Ha of vineyards in Catalonia (Spain) for phytoplasmas involved in grapevine yellows diseases.Molecular identification of phytoplasmas was performed from selected positive samples collected during different years of the epidemic.

Monitoring the presence of phytoplasma and plant material used for phytoplasma detection
In the Alt Empordà (Gerona), foci of FD were discovered by visual inspection of plots having more than 20% of their grapevine plants with symptoms of yellowing or reddening.This was done whenever possible by observing the plots from a high vantage point.Otherwise, the perimeters of the plots were surveyed to find symptomatic plants.Once one focus was found, plots at a distance of 500 m from the outer plants in the infected focus were inspected plant by plant.The rest of the plots in the area at risk (Alt Empordà) were examined at random.In the remaining vine-growing areas of Catalonia, samples expressing symptoms were selected.
Samples from 112 symptomatic grapevines were collected in the affected provinces during July and September for the seven year period of 1996-2002.

DNA extraction, PCR amplification and RFLP and sequence analyses
Leaf mid-vein tissues were subjected to DNA extraction using a combined method, a concentration of phytoplasmas with PGB (Ahrens and Seemüller, 1992) and a DNA extraction performed with the E.Z.N.A. ® Plant MiniPrep Kit (Omega Biotek), as described by Martín & Torres (2001).
PCR was performed with Ready-to-Go PCR Beads (Amersham Biosciences) as previously described (Martín & Torres, 2001).Samples lacking DNA were used in each experiment as negative controls; positive and reference controls were as described in Fig. 3. Amplification of a partial rDNA cistron (16S rDNA, 16S/23S rDNA spacer region, trnI and the 5' part of 23S rDNA representing ca.1800 bp) with phytoplasma-universal primer pairs P1 (Deng & Hiruki, 1991) and P7 (Schneider & al., 1995).This was followed by nested PCR reactions using the R16(V)F1 and R16(V)R1 primer pairs specific for group 16SrV and the R16(I)F1 and R16(I)R1 primer pairs specific for groups 16SrI and 16SrXII (Lee & al., 1994).Then selected grapevine samples that gave positive results with the R16(V) primers (Tab. 1) were further analysed using nested-PCR with primers 16R758f and M23SR 1804r (Martini & al., 1999) on the P1/P7 amplicon.Moreover, these samples were analysed by nested PCR with primers FD9f3 and FD9r2 on FD9f2/FD9r amplicons , which represent the entire SecY gene and a portion of the rpL15 gene (Angelini & al., 2001).In addition, PCR reactions using the rp(V)F1 and rpR1 primers gave a product representing part of the ribosomal protein rpS3 gene (Lee & al., 1998).About 200 ng of each positive product was each separately digested with TruI, TaqI, AluI and Tsp509I restriction enzymes (Fermentas, Vilnius, Lithuania) for at least 16 hours, following manufacturer instructions.
All PCR products that were sequenced were first cleaned using the E.Z.N.A. Clean kit (Omega Biotech).Fragments were sequenced using the Big Dye Terminator reaction kit (Applied Biosystems) in an ABI Prism 377 (Applied Biosystems).Sequencing  : 1353, 1078, 872, 603, 310, 281, 271, 234, 194, 118 and 72. of both strands was performed using the R16F2 and R16R2 (Lee & al., 1995) or P1 and P7 primers.Bioedit TM software was used to identify the consensus sequence from the two strands of each amplification product.

Monitoring of phytoplasma presence
In Alt Empordà (Gerona) the number of plots having more than 20% of their grapevine plants with phytoplasma symptoms was zero in 1998, starting from values of 29 in 1996, and 20 in 1997.From 1998 onwards, there has only been one outbreak, this one occurring in 2001 (Tab.2).There are several municipalities in which scattered plants affected by yellows were detected: Agullana, Biure, Cabanes, Cantallops, Espolla, Masarac, Pau, Peralada and St. Climent.Affected municipalities form a triangle whose base is at the Pyrenees (natural border with the South of France) and its opposite vertex towards the South of Spain (Fig. 4).
In the remaining Catalonian vineyards only isolated grapevines showed symptoms of phytoplasma disease.

Phytoplasma detection
Nested-PCR with the R16(V) and R16(I) primers showed that 45 samples over 112 analysed were positive for the presence of phytoplasma.The 20 samples that gave the PCR products of expected size using the R16(V) specific primers came from Gerona, while the 25 samples that gave PCR products of expected size with R16(I) specific primers were distributed between the four provinces.R16(V) primers indicated that these phytoplasmas were FD and were molecularly undistinguishable from each other and from reference strain FD88.In all the genomic fragments analysed, they were referable to phytoplasmas belonging to the ribosomal subgroup 16SrV-D and to rp subgroup E (Fig. 3) (Martini & al., 2002).No sequence differences were seen among the eight P1/P7 or R16F2/R2 amplicons (Tab.1).The BLAST search of sequence AJ548787 (1782 bp) from the Spanish isolate FD1487 showed highest similarity (99.9%) with the strain FD70 (AF176319) (Davis & Dally, 2001) from FD infected grapevines.The alignment of both sequences showed only two nucleotide differences, in positions 1425 and 1594 in the Spanish isolate FD1487.In agreement with the RFLP analysis of the ribosomal DNA fragment, the difference in nucleotide position 1425 corresponds with a putative restriction site for TaqI in FD1487 that is not present in FD70.The difference in position 1594 was due to a poliA (seven A in FD1487 and eight A in FD70).A complete 16S rDNA sequence of FD88 was not available, but AF458380, a partial 16S, 16S-23S intergenic spacer and 23S rDNA sequence of strain FD92 (=FD88) (Angelini & al., 2003), made it possible to compare 1020 nucleotide positions of AJ548787 (FD1487) with AF458380 (FD92=FD88).These sequences were 97.9% similar, however, both strains show the presence of a putative restriction site for TaqI at position 1425 that allows detection of identical profiles after RFLP analyses (Fig. 3).

BN identification
No significant sequence differences were seen among the P1/P7 amplicons of isolate BN2642 and the R16F2/R2 amplicons from isolates BN974, BN1536, BN2001 and BN 2002.The BLAST search of sequence AJ964960 (1684 bp) from Spanish isolate BN2642 showed highest similarity (99.5%) with sequence AF248959 from strain STOL, a member of the 16SrRNA RFLP subgroup XII-A (Davis & Dally, 2001).The alignment of both sequences showed only five nucleotide differences, in position 623, 714, 1007, 1013 and 1203 in isolate BN2642.

Discussion
During the monitoring of phytoplasma presence in the Catalonian provinces at risk (Barcelona, Gerona, Lérida and Tarragona), only grapevines at Alt Empordà (Gerona) were positive for disease caused by phytoplasmas from the 16SrV-D/elm yellows subgroup (flavescence dorée, FD).No phytoplasmas be-longing to this group were detected in the grapevines of the remaining plots of Gerona and of the other three Catalonian provinces where only phytoplasmas belonging to the 16SrXII-A/stolbur subgroup (Bois noir disease) were identified.
Bois noir disease is distributed in more viticultural areas, but with a low incidence level, and it is responsible for minor economic losses (observations of Servei Sanitat Vegetal, DARP).Moreover, flavescence dorée, a destructive phytoplasma that is subjected to quarantine regulations, is confined to a small part of the Alt Empordà near the French border.
The FD disease incidence in Spain has drastically decreased since 1996 when the first focus appeared.The restricted spread of FD could be a result of the effectiveness of the measures adopted (Rahola & al., 1997) such as the elimination of affected grapevines and the containment of S. titanus, even though the confluence of other natural factors that could have accounted for the decreased incidence can not be excluded.
The IRPCM Phytoplasma/Spiroplama Working Team-Phytoplasma Taxonomy Group ( 2004) established the basis to identify 'Candidatus Phytoplasma' species by the level of sequence identity of their 16S rRNA genes.The phytoplasma 16S rRNA sequences obtained from Spanish grapevines shows highest similarity with sequences that are related to two 'Ca.Phytoplasma' species that are not yet formally described but for which names were proposed at the Tenth International Congress of the International Organization of Mycoplasmology held in 1994 in Bordeaux, France.These names are 'Ca.Phytoplasma vitis' and 'Ca.Phytoplasma solani'.All sequences from samples with positive R16(V)F1/R1 amplicons are related to 'Ca.Phytoplasma vitis' of 16SrV/elm yellows group; sequences from samples with positive R16(I)F1/R1 amplicons are related to 'Ca.Phytoplasma solani' of 16SrXII/stolbur group.Both 'Ca.Phytoplasmas' reported here are incidental citations which do not constitute prior citations, according to rule 28b of the bacteriological code (Lapage & al., 1992).
Previous work has been conducted to define the subgroups of FD phytoplasmas.Davis & Dally (2001) classified two FD strains, based on 16S rRNA gene RFLP patterns, as members of two distinct subgroups: FD70 as 16SrV-C and X76560 as 16SrV-D.Genetic variability among FD phytoplasmas, based on RFLP analyses from three DNA fragments (partial rDNA operon, partial rps3 operon and partial SecY gene) was also reported by Martini & al. (1999Martini & al. ( , 2002) ) and Botti & Bertaccini (2003).Based on TaqI restriction site data, the latter study allowed the differentia-tion of two rRNA subgroups from grapevine samples, 16SrV-C and 16SrV-D, and allowed differentiation of seven FD phytoplasma variants belonging to the 16SrV-C (including strain FD70) and two FD variants belonging to the 16SrV-D (including strain FD88) subgroups.The subgroups defined by Davis and Dally (2001) are not equivalent to those described by Martini & al. (1999Martini & al. ( , 2002)), since they were based on RFLP (TaqI) patterns of the 16S rDNA cistron in which only two strains belonging to subgroup 16SrV-C were compared.
Direct comparison or phylogenetic analysis of 16S rDNA provides a good tool to define phytoplasma groups and to identify 'Ca.Phytoplasma' taxa (IR-PCM, 2004), but the high levels of similarity among different 16SrV phytoplasmas using this DNA fragment preclude its use in subgroup classification (Davis & Dally, 2001).It seems that the rps3 gene shows more variability and can be employed to differentiate closely related FD strains (Martini & al., 2002).

Fig. 1 .
Fig. 1.Symptoms of decline and reddening on a red wine grapevine cultivar.

Fig. 2 .
Fig. 2. Symptoms of leaf rolling and yellowing on a white wine grapevine cultivar.

Table 1 .
Description of selected flavescence dorée phytoplasma isolates analysed by RFLP.Genbank accession numbers from sequences of some isolates are shown. *