In this study quantitative real-time PCR was used to follow the seasonal changes of flavescence dorée phytoplasma (FDp) titre in grapevines of cv. Modra frankinja (syn. Blaufränkisch) and cv. Refošk (syn. Refosco'd'Istria) from two vineyards located in climatically different vine-growing regions of Slovenia. Besides its known presence in the leaf veins, FDp was also detected in flowers, berry tissues and tendrils. In plants with high concentrations of FDp in tissues with symptoms, phytoplasma was also detected in symptomless tissues. A trend of decreasing FDp titre in all examined symptomless tissues from June to July and an increasing one throughout the growing season in tissues with symptoms was recorded. Accordingly, FDp was present in detectable amounts in flowers, petioles and veins of almost all infected plants in the late spring, and was detected in all examined tissue types in summer, with the highest titre in berries in August. The study showed that in the absence of plant health measurements an FDp infection may spread exponentially by a factor of 40 per year.
At least 10 phytoplasma ribosomal subgroups have been associated with grapevine yellows (GY) diseases, with a great economic impact on viticulture (Constable, 2010). Different grapevine-infecting phytoplasmas cause nearly identical symptoms, which are indistinguishable by visual inspection, but the epidemiology associated with each phytoplasma species is different (Constable, 2010; Belli et al., 2011). In Europe the main phytoplasmas associated with GY are the flavescence dorée (FD) and bois noir (BN) phytoplasmas. Based on 16S rDNA sequence similarity FD belongs to 16SrV, subgroups C and D (Lee et al., 2004) and BN to the 16SrXII-A (Lee et al., 2007) ribosomal group. The occurrence of FD has been catalogued as catastrophic in France and Italy. In the middle of the 1950s FD occurred in Armagnac and rapidly reached epidemic status, spreading up to 10 km a year. Up to 80% of vines were diseased and yield losses of 20–30%, at times as high as 80%, were reported (reviewed by Magarey, 1986). Later reports of the disease in northern Italy also described strong yield reduction and a disease incidence of 50% or more (reviewed by Belli et al., 2011). Currently, FD is widespread in many vine-growing regions of France and Italy, outbreaks have occurred in Portugal, Serbia and Slovenia, and a few occurrences have been recorded in Spain, Switzerland and Austria (PQR database, EPPO). Because of its epidemic potential, the FD phytoplasma (FDp) is listed in the EU2000/29 Council Directive on Harmful Organisms and the EPPO A2 quarantine list of pests, and the destruction of diseased stocks, plants showing symptoms and surrounding plants is mandatory. Accordingly, the quarantine status of FDp limits studies on its interaction with the host grapevine and with their insect vector Scaphoideus titanus. The lack of information on grapevine–FDp interactions is additionally associated with the current general inability to cultivate phytoplasmas in vitro.
Although not specifically proven for FDp, it has been shown that other phytoplasmas are unevenly distributed, especially in woody hosts, and their titre varies according to the season and plant organ (Galetto & Marzachi, 2010). Using ELISA for FDp detection in several grapevine tissues in summer months, Caudwell & Kuszala (1992) determined leaf veins to be the sampling tissue of choice, and they are currently used as sampling tissues in molecular diagnostics (Angelini et al., 2001, 2006, 2007; Bianco et al., 2004; Galetto et al., 2005; Hren et al., 2007). However, there is no study indicating that leaf veins are the only good source of FDp throughout the whole growing season.
The aim of this work was to determine the distribution of FDp in different grapevine aerial tissues at several time points of two successive growing seasons in two FDp-infected vineyards from vine-growing regions with different climatic conditions. In order to follow the hypothesized FDp seasonal and tissue titre changes, quantitative real-time PCR (qPCR) was used. The potential contribution of this study to official FDp diagnostic schemes is discussed.
Materials and methods
Locations and sampling
The study was carried out in growing seasons 2010 and 2011 in two distant production vineyards in Slovenia – a vineyard of cv. Modra frankinja (syn. Blaufränkisch) located in the southeastern Dolenjska region (45°47′N, 15°4′E) (referred to as vineyard A) and one of cv. Refošk (syn. Refosco'd'Istria) in the southwestern Primorska region (45°31′N, 13°40′E) (referred to as vineyard B). The two regions have different climate conditions, vineyard A being in a continental climate with snowy winters and hot summers, while vineyard B has a mild Mediterranean climate. Both vineyards have single Guyot vine training. The incidence of FDp infection in vineyards A and B was moderate and very high, respectively.
In vineyard A a short row was selected, which included four plants with symptoms with confirmed presence of FDp in July 2010, one plant with atypical phytoplasma symptoms and 11 symptomless plants in which FDp was not detected by qPCR. The whole row was protected with a quarantine net to avoid obligatory uprooting of infected plants and thus allow their growth in subsequent seasons. The use of the net was approved by the Phytosanitary Administration of the Republic of Slovenia. The plants inside the net were subjected to the same antifungal treatments and pruning as the plants outside the net but were more frequently treated against the FDp vector S. titanus. After the initial sampling in July 2010, additional samplings were performed in August and October 2010 and at the time of pruning in March 2011, followed by sampling once a month during the growing season up to August 2011.
Vineyard B neighboured a vineyard which was completely uprooted in 2010 because of an extremely high FDp infection rate. In 2010 there was no spraying against the FDp vector S. titanus in the region. Several presumably infected plants from the first four rows were selected for the study. The first sampling of three plants was done at the time of pruning in February 2011; in May and June 10 plants were sampled, while in July and August six plants were sampled.
At every sampling time various tissues were sampled according to their availability during the growing season (Table 1). Generally, a whole set of tissues from the individual plant were collected from one shoot. Where there were both shoots with and shoots without symptoms on an individual plant, sets of tissues from both shoot types were sampled separately. FDp presence was determined in petioles, the three main veins and laminas cut separately with sterile scalpels from fully developed leaves. Berry cluster peduncles were also collected, using a sterile scalpel. Tendrils, flower clusters from 1–3 flower bunches and berries from different parts of the berry cluster with berry pedicels still attached were pulled by hand. From approximately 10 berries, berry pedicels were then separated and pooled. Remaining berries were divided into two groups. One group represented whole-berry samples, and from the other group berry skin was separated from berry flesh by squeezing the flesh out, thus forming separate berry skin and berry flesh samples. Scrapings of the secondary phloem-rich bark were collected from 1-, 2- and 3-year-old canes. All tissue samples were frozen and stored at −80°C until further analysis. Overall, over 520 tissue samples were collected and analysed.
Table 1. Reliability of flavescence dorée phytoplasma (FDp) detection in individual grapevine tissues with (S) and without symptoms (AS) in two vineyards in Slovenia. Data are the number of plants in which FDp was detected out of the total number of tested infected plants; shading represents percentage of positive samples: , 0%; , 1–59%; , 60–99%; , 100%
Empty squares, not sampled.
In July 2010, berries and berry tissues with and without symptoms were not sampled separately.
Sampled tissues were homogenized to powder in liquid nitrogen using a pestle and mortar. DNA extraction was based on magnetic beads using a KingFisher mL machine (Thermo Scientific) and QuickPick™ SML Plant DNA Kit (Bio-Nobile), and was carried out according to Pirc et al. (2009) with minor modifications. Tissue powder (100–200 mg) was mixed with lysis buffer in the ratio 1 mg per 3 mL and with 25 μL proteinase K. After 30 min incubation at 65°C and centrifugation for 1 min at 6000 g, 220 μL extract was used for DNA isolation. At the end of the isolation procedure DNA was eluted in 200 μL water.
FDp detection and quantification
Detection of FDp and the plant 18S rRNA gene as an endogenous control was done with qPCR performed on the Roche LightCycler 480 (Roche) in 384-well plates, using primers and TaqMan MGB probes described by Hren et al. (2007). qPCR was performed in a final reaction volume of 10 μL containing 2 μL sample DNA, 1× Maxima Probe qPCR master mix (Fermentas) and 900 nm primers plus 250 nm probe for FDp or 1× primer probe mix (AB-Applied Biosystems) for 18S rDNA. Universal cycling conditions (2 min at 50°C, 10 min at 95°C, followed by 45 cycles of 15 s at 95°C and 1 min at 60°C) were used. Each DNA sample was tested for FDp and plant 18S rDNA. All reactions were performed in two replicate wells at two, 5-fold different dilutions. For quantitative analysis a reference sample (sample containing FDp and 18S rDNA) was used for standard calibration curve preparation for both amplicons and was placed on each reaction plate. LightCycler 480 Software (Roche, Germany) was used for fluorescence acquisition and automatic Cq (quantification cycle) calculation.
For a relative quantification of FDp titre, relative FDp amount was calculated using a standard curve quantification approach (Mehle et al., 2012). A standard calibration curve was calculated for each amplicon from serial dilutions of the reference sample and the limit of FDp quantification was defined at Cq value 33. Samples that had Cq values below the limit of quantification or did not fulfil the quality criteria were discarded. For samples having an amount of FDp DNA below the limit of quantification and above the limit of detection, a relative FDp amount corresponding to the calculated copy number at the limit of quantification on the standard calibration curve was assigned. Similarly, to enable further calculations, copy numbers 25-times lower than the limit of quantification were also assigned to the negative samples. Normalized relative amounts of FDp DNA were log2-transformed and then compared between the samples.
A parallel molecular study of the FDp isolates involved in this analysis revealed that all sampled plants in both vineyards were infected with isolates belonging to the 16SrV-D subgroup (Mehle et al., 2011).
Each grapevine plant included in the study was also tested for the presence of bois noir phytoplasma (BNp) according to the detection system developed by Hren et al. (2007). Testing revealed no BNp present in the selected FDp-infected plants.
Symptom development during the 2011 growing season
At the time of pruning in the beginning of March, the canes of FDp-infected plants had poorly matured wood. As a consequence, by May these plants showed reduced growth and fewer flower clusters (Fig. S1). The first symptoms of leaf reddening developed just after flowering in June. At the same time some berry clusters began to fall off. Symptoms on leaves were more pronounced in July. In addition to more extensive reddening, leaves also started to curl downwards. In August, leaf reddening began to develop into a purplish colour. The number of berry clusters on the plants with symptoms was lower than on symptomless plants, and berry withering was observed on the former. Moreover, some severely infected shoots were without grape clusters. However, in August in vineyard A two plants did not show symptoms on newly developed leaves.
FD disease spread in vineyard B
In August 2010, 26 plants limited to the first four rows of vineyard B, out of approximately 1800 plants in the whole vineyard (i.e. 1·4%; Fig. S2), displayed typical symptoms of FDp infection (Fig. S1), and FDp was confirmed by the official survey. In 2011, the number of plants with symptoms in the entire vineyard was more than 1000 (Fig. S2). This represents nearly 60% of all plants or a 40-fold increase from 2010 to 2011. Because of this extensive infection the entire vineyard was uprooted just after harvest.
FDp detection in a quarantine net in vineyard A
In summer 2010, the presence of FDp was confirmed in all tissue types on shoots with symptoms, with the highest FDp DNA titre detected in August in berry skin (Table 2). The FDp titre estimated by qPCR was higher in most tissues in August than in July (Table 2). In August the relative FDp titre was higher in berries from shoots with symptoms than in leaf veins (P =0·034) and tendrils (P =0·00038; Table 2). It was similarly high in all individual berry tissues. The only exception was a significantly lower amount of FDp DNA detected in berry cluster peduncles (P =0·0023; Table 2). FDp was also present at detectable concentrations in berry tissues on the symptomless shoots of a plant without berries on its shoots with symptoms (Tables 1 & 2), but its average titre was 103- to 104-fold lower (P =0·000046) than that of berry tissues on shoots with symptoms (Table 2).
Table 2. Relative quantification of flavescence dorée phytoplasma (FDp) in different grapevine tissues with (S) and without symptoms (AS) in 2010 and 2011 in vineyard A. Estimated relative FDp copy number is normalized to 18S rRNA copy number. In order to estimate the increase/decrease in relative FDp amount, FDp negative samples were also assigned a relative FDp amount corresponding to the calculated amount on the standard calibration curve. Results are the means of log2-transformed copy numbers of FDp ± SE. Numbers of tested samples are shown in Table 1
NA, tissue sample not available.
In July 2010, berry tissues with and without symptoms were not sampled separately.
In canes, FDp was barely detected, except in the secondary phloem of a 1-year-old cane with symptoms in July, and at the end of the growing season in October 2010 when FDp was present in the secondary phloem of 2-year-old canes in two out of four plants (Table 1). In March 2011, the FDp titre was below the limit of detection in 3-year-old canes, in which FDp had been confirmed in the autumn (Table 1).
At flowering and until June, all tissues were symptomless and FDp was not detected. The first unambiguous proof of FDp presence in 2011 was in June, when symptoms developed on one out of four previously infected plants. In tissues with symptoms from this plant, low amounts of FDp DNA were detected in leaf veins, leaf petioles and tendrils (Tables 1 & 2). However, in a few plants FDp could also be detected in symptomless tissues of flowers, leaf veins, leaf petioles and tendrils (Tables 1 & 2).
In July, when symptoms appeared in all four infected plants, the presence of FDp was confirmed in leaf veins and petioles with symptoms from all of them (Table 1). FDp was also detected in whole berries and berry pedicels of all plants with berries on the shoots with symptoms. In accordance with milder symptoms in August 2011, the average relative amount of FDp in berry tissues on shoots with symptoms was lower than in August 2010 (P =0·0016; Table 2, Fig. 1a). By August 2011, only two plants retained symptoms on leaves, and FDp was detected in leaf and berry tissues from shoots with symptoms of these plants (Table 1), with the highest average titre in leaf veins and berry pedicels (Table 2).
In general, the average FDp titre steadily increased in tissues with symptoms from June to August in both growing seasons (Table 2, Fig. 1a). However, the differences were for the most part not statistically significant because of either high variability or low number of plants (Fig. 1). Statistical significance was found for the increase of FDp titre from July to August 2010 in leaf veins with symptoms (P =0·0039) and laminas (P =0·0055).
FDp detection in vineyard B
Flavescence dorée phytoplasma was detected in cane secondary phloem samples from all three analysed plants at the time of pruning in February 2011 (Table 1). Again in May, before symptom appearance, it was confirmed in symptomless leaf tissues, flowers and tendrils (Table 3) in the majority of analysed plants (Table 1). At the time of symptom development in June 2011, FDp was shown in the midribs of all examined shoots with and without symptoms, in all leaf petioles with symptoms and in 83% of symptomless leaf petioles, in 80% of flowers on symptomless shoots, and in all tendrils with symptoms and 83% of symptomless tendrils (Table 1). It was also confirmed in a high percentage of analysed canes (Table 1). Therefore, detection of FDp in symptomless tissues was relatively reliable, although the variability between plants was high (Table 3). The FDp titre in symptomless shoots in June was highest in berry pedicels at the time when berries had just started to develop (Table 3) and was comparable to that in leaf veins with symptoms (Table 3). It is noteworthy that at this time the average FDp titre in berries on symptomless shoots was estimated to be 304-fold higher than in berries on shoots with symptoms (Table 3, Fig. 1). It is also striking that from June to July the estimated average FDp titre in symptomless berries decreased 207-fold (P =0·03) and only slightly increased again in August (Table 3). On the other hand, it substantially increased (105-fold) from June to July in berries with symptoms (Table 3, Fig. 1). As in berries, a drop in FDp titre from June to July and a minor rise in August was also detected in other symptomless tissues (Table 3). An interesting observation was that the titre of FDp in leaf tissues with symptoms remained relatively stable throughout the season (Table 3, Fig. 1).
Table 3. Relative quantification of flavescence dorée phytoplasma (FDp) in different grapevine tissues with (S) and without symptoms (AS) in 2011 in vineyard B. Estimated relative FDp copy number is normalized to 18S rRNA copy number. In order to estimate the increase/decrease in relative FDp amount FDp negative samples were also assigned a relative FDp amount corresponding to the calculated amount on the standard calibration curve. Results are the means of log2-transformed copy numbers of FDp ± SE. Numbers of tested samples are shown in Table 1
In this work seasonal distribution of FDp was systematically followed in grapevine plants in two vineyards located in two regions with different climatic conditions and with different FDp infection rates, with the infection rate being very high in vineyard B. As well as being affected by location and cultivar, high infection rates were probably related to the presence of the main FDp vector, S. titanus, which was not controlled in 2010 in vineyard B. The estimated spread of FD disease in vineyard B was about 40-fold from 2010 to 2011, a rate even faster than that predicted by Steffek et al. (2007) in their pest risk analysis.
The average FDp titre in infected plants inside the quarantine net in vineyard A slightly decreased from one growing season to another. In some plants leaf symptoms were not even observed in leaves that developed later in the season and were associated with undetectable amounts of FDp. This result might be related to the absence of the vector S. titanus because of the protective net and extensive spraying against the vector. It was suggested previously that recovery occurs with time in the absence of reinoculation (Caudwell, 1961, 1964).
FDp was detected in all tested tissue types. Although its presence in grapevine flowers was demonstrated for the first time in this study, several other phytoplasma species have been detected in flowers of various plants (Siddique et al., 1998; Azadvar et al., 2011; Oropeza et al., 2011; Su et al., 2011).
The detection of FDp in symptomless tissues early in the growing season was quite successful. This study showed that in June the detection of FDp in symptomless tissues was positively correlated with its overall titre in the plant. The probability of FDp detection in symptomless shoots was lower in vineyard A than vineyard B, where the average FDp titre in shoots with symptoms was higher. In a study of FDp-infected plants of cv. Barbera, no FDp was detected before July (Bulgari et al., 2011). Whether the observed changes in FDp titre are related to cultivar-specific susceptibility or different vineyard climate and growing conditions is currently not known.
The increasing titre of FDp during the growing season has been reported in grapevine leaves (Bulgari et al., 2011). The present study detected a similar trend not only in leaves with symptoms, but also in berries on shoots with symptoms, where a very sharp increase in FDp titre was detected from June to July followed by a steadier increase until the end of the growing season.
Although this is the first systematic study of FDp presence in grapevine berries, its occurrence in this tissue is not unexpected. Mycoplasma-like structures have been observed in the phloem cells of peduncles and pedicels of phytoplasma-infected grapevine plants showing premature berry dehydration disorder (Matus et al., 2008). There is accumulating evidence that phytoplasma infection interferes with host plant sugar metabolism and that levels of reducing sugars and sucrose are generally higher in the source leaves of plants that are infected with various phytoplasma species (Hren et al., 2009 and references therein). Therefore, it may be possible that phytoplasmas invade tissues by growing along sieve elements, perhaps in response to sugar gradients, and either multiply or die off according to the availability of nutrients. Interestingly, there was a decrease in FDp titre in all symptomless tissues from June to July in vineyard B. This was also observed in a few plants with symptomless samples in vineyard A. At the moment it is not known if the higher titre of FDp in tissues with an increased sugar concentration is related to FDp-induced transport/synthesis of metabolites, specifically sugars, to/within tissues that consequently express symptoms, or if this increase is a result of plant responses to the infection.
The results of this study may result in more reliable and accurate diagnostics of FDp, allowing detection very early in the season, possibly before symptom appearance. They confirmed the findings of previous studies on different grapevine yellows phytoplasma species (Del Serrone & Barba, 1996; Gibb et al., 1999; Constable et al., 2003; Terlizzi & Credi, 2007), which determined that phytoplasma detection is unambiguous when the first symptoms appear and continues through the summer months. However, in addition to the widely used leaf veins with symptoms as a phytoplasma source, berries from shoots with symptoms are a similarly reliable tissue for detection, with high FDp titre. If the infection pressure in a vineyard is very high, the presence of FDp may be detected in cane samples as early as February and as late as May in symptomless leaf tissues, flowers and tendrils. Flowers appear to be the tissue of choice for a spring FDp diagnosis, and for even higher reliability they may be used in combination with leaf veins. Although in routine diagnosis, molecular testing is usually just a confirmation of visual diagnosis, the testing procedure described in this study may be useful for particular aims. For example, if a vineyard experienced FDp infection in the previous season, but not all plants showed symptoms of FD and so were not discarded, early testing in the next season would allow additional infected plants to be destroyed in order to prevent further spread of the infection.
This research was supported by the EU project 262032VITISENS (in which some SMEs participate), and by the research grant ARRS-1000-09-310032 from the Slovenian Research Agency. Samples were taken during the official survey of FD in Slovenia by the Slovenian Phytosanitary Administration. All parties have agreed to the publication of the results. The authors thank the Slovenian Phytosanitary Administration for allowing the construction of the quarantine net in vineyard A and Mr Ivan Krakar for maintaining the plants inside it. The authors also thank Nataša Tešić, Jernej Pavšič and Špela Zupančič for their help with sampling and sample processing, and Dr Aleš Kladnik for data presentation. The authors are grateful to Drs Neil Boonham and Elizabeth Covington for their critical reading of the manuscript.