Introduction

Grapevine is an ancient culture that currently constitutes the most economically important fruit species worldwide. Diverse pathogens attack this species causing physiological disorders that finally lead to yield losses. Among them, viral diseases can have a serious impact on grapevine productivity and fruit quality. Viral infections are usually disseminated through biological vectors and vegetative propagation of virus-infected material. These infections affect the vegetative organs inducing foliar deformations, alterations in leaf color, and, in some cases, graft rejection (Martelli 1993). Severe infections also reduce berry setting and cause irregular and delayed ripening (Martelli 1993). Among the more than 40 different viruses that infect grapevines (Martelli 1993), the leaf roll-associated closterovirus-3 (GLRaV-3) is one of the most wide spread. GLRaV-3 belongs to the Ampelovirus genus of Closteroviridae and is a positive-sense ssRNA virus transmitted by grafting and mealy bugs (Ling et al. 2004). Diseases associated with this virus affect fruit quality by delaying ripening, diminishing sugar content, and, in red cultivars, reducing color (Borgo and Angelini 2002).

Grapevines develop an active resistance response against pathogens like fungi and bacteria, and some resistance genes have been identified (Aziz et al. 2003; Vivier and Pretorius 2002; Di Gaspero and Cipriani 2002). To date, no active resistance response against viral diseases has been found in grapes, and viruses establish a compatible interaction causing disease spread through all plant tissues. However, it has been shown that diverse susceptible hosts are not passive against a pathogen and can set up a defense response that could be less intense and not enough to stop viral replication and dissemination (O’Donnell et al. 2003; Ehrenfeld et al. 2005). Visible plant symptoms represent the sum of molecular, cellular, and physiological changes induced by viruses that are part of the host defense response. In compatible interactions, multiple changes in gene expression underlie the disease symptom development and also control the levels of virus in infected tissues (Maule et al. 2002).

The DNA microarray techniques that permit the simultaneous analysis of a large number of genes are now contributing to understanding the plant–pathogen relationship (Wan et al. 2002). Studies of compatible interactions in the model plant Arabidopsis thaliana have shown that different viruses induce plant gene expression changes on a wide array of cellular processes (Golem and Culver 2003; Whitham et al. 2003). These changes include the induction of genes associated with defense and stress, signal transduction, and metabolism (Whitham et al. 2003). On the other hand, a shut-off phenomenon in host gene expression has been observed during compatible interactions affecting different cellular process (Maule et al. 2002; Golem and Culver 2003, Whitham et al. 2003).

With the aim to investigate the effect of viral infections on the grapevine, transcript profiles of the red wine cultivars Carménère and Cabernet-Sauvignon naturally infected with GLRaV-3 were compared with virus-free grapevine plants. These two cultivars were chosen because both are susceptible to leaf roll disease, Cabernet-Sauvignon is one of the most economically valuable grapevine in the world, whereas Carménère has been acquiring relevance in Chile, which is reflected by the notable increase in the area cultivated during the past years. The application of genomic strategies to the functional study of grapevine is now possible due to the efforts in obtaining expressed sequenced tags and transcript analysis in these species (Goes da Silva et al. 2005). In the present study, we have used the recently developed Vitis vinifera GeneChip® microarray that contains 14,000 V. vinifera transcripts and 1,700 transcripts from other Vitis species. We describe in this paper the identification of numerous genes that are induced or repressed during viral infection in grapevine associated with two specific cultivars. The study contributes to understanding the changes triggered by viruses on grapevine physiology and to developing some proposals for mechanisms that infected grapevines utilize to contend the viral pathogen.

Materials and methods

Plant material

Healthy and GLRaV-3-infected grapevine plants were obtained from the nursery of Agronomy Faculty, P. Universidad Católica de Chile. Nine viruses that have a high incidence in Chile were screened by ELISA tests (Agritest) following manufacturer conditions: Grapevine Virus A (GVA), Grapevine virus B (GVB), Grapevine Fanleaf Virus (GFLV), Grapevine Fleck Virus (GFkV), Tomato Ringspot Virus (ToRSV), and Grapevine Leaf-Roll-Associated Viruses (GLRaV) 1, 2, 3, and 7. Healthy and virus-infected plants were maintained separately in greenhouses with controlled growing conditions. Medial leaves (approximately 8 cm diameter) from 12 plants were collected during growing season, to assure collecting leaves of about the same age. These included healthy Carménère and Cabernet-Sauvignon plants used as control and GLRaV-3 infected Carménère and Cabernet-Sauvignon plants used as infected tissue.

RNA extraction and RT-PCR

Total RNA was extracted from GLRaV-3-infected and healthy Vitis vinifera cv. Carménère and Cabernet-Sauvignon leaves following the Tris–LiCl method modified by Goes da Silva et al. (2005). Briefly, samples were extracted with homogenization buffer (200 mM Tris–HCl, pH 8.5, 1.5% (w/v) lithium dodecylsulfate, 300 mM LiCl, 10 mM sodium EDTA, 1% (w/v) sodium deoxycholate, 1% (v/v) Nonidet P-40, 2 mM aurintricarboxylic acid, 20 mM dithiothreitol, 10 mM thiourea and 2% (w/v) polyvinylpolypyrrolidone). After RNA precipitation with sodium acetate and isopropanol, samples were extracted with 25:24:1 phenol:chloroform:isoamyl alcohol and 24:1 chloroform:isoamyl alcohol before performing LiCl precipitations. Before RT-PCR reactions, samples were treated with DNase (Invitrogen). Then, first-strand cDNA was synthesized with random hexamers and reverse transcriptase (Stratagene). An aliquot of 1 μl was used as template for PCR. For GLRaV-3 confirmation, the primers C50 and H59 that amplify 945 bp of coat protein (Minafra and Hadidi 1994) were used.

Microarray hybridization

Six Carménère (three healthy and three infected) and six Cabernet-Sauvignon (three healthy and three infected) plants constituted the biological samples for expression analysis. RNA quality and concentration was measured using an Agilent 2100 bioanalyzer charging 2 μl of each sample in a RNA 600 Nano Lab Chip. Samples were processed following the GeneChip® Expression 3′-Amplification Reagents One-cycle cDNA synthesis kit instructions (Affymetrix Inc, Santa Clara, CA, USA). Single-stranded cDNA was synthesized with 6.5 μg of RNA of each sample, and oligo-dT-T7Promoter Primers using the Superscript II (Invitrogen, USA). Then, double-stranded cDNA was synthesized and used as template to generate biotin-targeted cRNA following manufacturer’s specifications. Fifteen micrograms of the biotin labeled cRNA was fragmented to strands between 35 and 200 bases in length. The fragmented cRNA (10 μg) was hybridized on a Vitis vinifera GeneChip® genome array (http://www.affymetrix.com/products/arrays/specific/vitis.affx) using standard procedures (45°C for 16 h). The chip is composed of approximately 14,000 V. vinifera probe sets and 1,700 probe sets of other Vitis species and hybrids, allowing the exploration of about 30–50% of the grape transcriptome. Chip probes were designed from public available databases that contain sequences from cDNA libraries of different V. vinifera cultivars, mainly Cabernet-Sauvignon and Chardonnay, among others. Libraries include sequences of several developmental stages and tissues, such as berry, leaf, flowers, and roots, and also contain ESTs from different conditions or treatments, comprising biotic or abiotic stresses. All sequences are accessible at Affymetrix web site (http://www.affymetrix.com/products/arrays/specific/vitis.affx) and in grapevine databases (http://www.barleybase.org and http://www.plexdb.org). Each transcript is represented by 16 pairs of 25-mer probes synthesized in situ on the array. The arrays were washed and then stained in a Fluidics Station 450. Scanning was carried out with the GeneChip® scanner 300 and image analysis was performed using GeneChip® Operating Software.

GeneChip® data processing and analysis

GeneChip® arrays data were first assessed using a set of standard quality control steps described in the Affymetrix manual GeneChip ® Expression Analysis: Data Analysis Fundamentals (http://www.affymetrix.com/products/arrays/specific/vitis.affx). Calls of all three spike-in controls BioC, BioD, and Cre were present, and their intensity values increased from BioC to Cre, as expected. Average background values ranged from 29 to 106, and only two arrays had background values greater than 100, the suggested limit. The range of scaling factors was 1.1 to 5.9, with two arrays exhibiting scaling factors greater than the optional threshold of 5. Noise values, as measured by raw Q-values, varied between 0.8 and 3.1. Digestion curves displaying trends in RNA degradation between the 5′ end and the 3′ end in each probe set were also inspected, and all 12 proved very similar.

Arrays data were processed and normalized by RMA (Robust Multi-Array Average) (Irizarry et al. 2003) using the R package affy (Gautier et al. 2004). Specifically, expression values were computed from raw CEL files by first applying the RMA model of probe-specific correction of PM (perfect match) probes. These corrected probe values were normalized via quantile normalization, and a median polish was applied to compute one expression measure from all probe values. Resulting RMA expression values were log2-transformed. Both Pearson correlation and Spearman rank coefficients were computed on the RMA expression values (log2-transformed) for each set of biological replicates. Spearman coefficients ranged from 0.903 to 0.988 and Pearson coefficients ranged between 0.913 and 0.989.

To determine which genes were differentially expressed between each infected cultivar and its corresponding control state, an ANOVA was performed on the RMA expression values (Kerr et al. 2000). The following model was used for this analysis: \(y_{{ijk}} = C_{i} + T_{j} + {\left( {CT} \right)}_{{ij}} + \varepsilon _{{ijk}}\), where y ijk denotes the log2 signal measured for cultivar i, treatment j, and biological replicate k, with 1≤i≤2, 1≤j≤2, and 1≤k≤3. The terms C i and T j measure the effect of the cultivar and treatment, respectively, and the interaction term (CT) ij accounts for the interaction between cultivar and treatment.

An ANOVA was performed on each gene using the linear model above and three contrasts based on differences between cultivars, and each cultivar’s stress and control state. The R package limma was used for ANOVA methods (http://www.bioconductor.org/repository/devel/vignette/affy.pdf) A multiple-testing correction (Benjamini and Hochberg 1995) was applied to the p values of the F statistics to adjust the false discovery rate. Genes with adjusted F statistic p values <0.05 were extracted for further analysis.

Semiquantitative RT-PCR

For semi-quantitative RT-PCR (sqRT-PCR), 2.5 μg of total RNA DNase-treated were used to synthesize cDNA, as was previously described. PCR reactions were done with 1 μl of cDNA samples. A master mix was prepared for each gene and all sqRT-PCR were performed in biological triplicate, in an Eppendorf Mastercycler Gradient. To establish the kinetic of amplification of each gene, PCR reactions covering from 19 to 44 cycles were performed. For all genes evaluated, the logarithmic amplification was observed with 31 cycles and this point was considered to compare gene expression. The following thermal cycle conditions were used: 94°C for 3 min, followed by 94°C for 50 s, 56°C for 50 s and 72°C for 3 min for 31 cycles. Amplified grapevine genes were lipid transfer protein (TC: 46011), harpin-induced 1 (TC: 40876), and tropinone reductase (TC: 40352).

Specific primers for the three genes (LTP-Vv-5′:CCACACCATATGCAGAGGAG and LTP-Vv-3′:AATGGAGCTGAGCACATGTG; HIN1-Vv-5′:ATGCCATCACAGCAACACA and HIN1-Vv-3′:TCAGGGCCTCATCTTCTCA; TROP-Vv-5′:GCTTCCTGCTCCAGATGC and TROP-Vv-3′:TGCCACCGTACACACCTG) were designed based on the sequences of the TC numbers (http://www.tigrblast.tigr.org/tgi/). All PCR products have an average length of 300 bp. Relative expression levels of each sample were normalized to the expression level of GPDH, TC: 44984 (primers GPDH-Vv-5′:CAGGATGCCATGTGGACAA and GPDH-Vv-3′:GTGTTGCCTTCATTGAATGG), which was expressed at a constant level in our experimental conditions. Quantification was done using Scion Image beta 4.0.2 software (Scion Corporation).

Results

Viral infection induces changes in grapevine transcript profiling in a wide spectrum of biological functions

Initially, virus-free and naturally GLRaV-3 infected grapevine plants were selected because the transmission of viral diseases to grapevine plants in controlled experimental conditions has turned out to be difficult to implement. The GLRaV-3 infection was confirmed by RT-PCR. The group of virus-free plants was consistently negative for all the grapevine viruses. Changes in gene expression in virus-infected grapevine plants were addressed in the cultivars Carménère and Cabernet-Sauvignon. In each case, three independent biological replicates were used and all hybridizations showed a similar distribution of expression values. After data normalization, medians distribution was uniform, reflecting that expression values were corrected for technical variances (Fig. 1). The percent of probe set that showed hybridization on the chips (percent call rate) ranged from 58 to 72% across all arrays, indicating that about 9,000–11,300 genes were expressed at detectable level in leaf tissues.

Fig. 1
figure 1

Boxplots of normalized perfect match values. The figure shows data distribution of log2 perfect match (PM) values for all arrays after processing and normalization. Arrays were hybridized with the following samples: RNA from Carménère and Cabernet-Sauvignon control non-infected plants (CA-Ctrl and CS-Ctrl, respectively); RNA from Carménère and Cabernet-Sauvignon infected with GLRaV-3 (CA-Inf and CS-Inf, respectively). Numbers 1 to 3 represent biological independent replicates. Log2 (PM) values between the 25th and the 75th percentiles are grouped into each box. Upper and lower dotted lines denote values greater and smaller than the 90th and the 10th percentiles

Full size image

ANOVA showed 492 genes in the Carménère cultivar, and 1,216 genes in Cabernet-Sauvignon with significant differences between control and the infected state (Tables S1 and S2, respectively). Genes with absolute log-transformed differential expression values ≥1 were extracted for further inspection, resulting in 341 Carménère genes and 692 Cabernet-Sauvignon genes.

The assignment of biological roles permits an overview of the metabolic changes occurring during viral disease in grapevines (Fig. 2). This analysis was done considering the sequences of genes represented on the array and their annotation (Gene Ontology Consortium). The wide range of cellular processes affected by viral infection in both cultivars includes mainly functions related with protein turnover, transport, metabolism, transcription, cell defense, and an important number of genes belonging to the biological process unknown category.

Fig. 2
figure 2

Biological functions involved in grapevine response to viral infection. The figure shows the percentages of genes affected by viral disease distributed into the different biological functions for Carménère (a) and Cabernet-Sauvignon (b) cultivars. Only genes with differential expression of at least twofold were graphed, corresponding to 341 Carménère and 692 Cabernet-Sauvignon genes. To perform the biological role assignment, grapevine sequences (TCs from TIGR Grape Gene Index, (http://www.tigr.org/tdb/tgi), matching to probe sets on the array were aligned to Arabidopsis genome (WU-BLAST 2.0) and then all loci identifiers were used to obtain the GO annotation and functional characterization (http://www.geneontology.org)

Full size image

Cellular changes in grapevine leaves infected with virus involved both induction and repression of gene expression (Fig. 3). In Carménère, 52 genes were induced and 289 were repressed in viral infected tissues. A more detailed analysis of these genes showed that some functional categories are exclusively represented by repressed genes, such as those involved in biosynthesis and cell cycle. In Cabernet-Sauvignon, 299 genes were induced and 393 were repressed in viral infected tissues. Overall, an important proportion of genes associated with unknown biological processes appear repressed. In contrast, other biological processes are predominantly induced, such as the genes involved in energy, transport, and signal transduction.

Fig. 3
figure 3

Distribution of up- and down-regulated genes during viral infection into functional categories. The distribution of genes induced or repressed by viral disease in grapevine cultivars Carménère (a) and Cabernet-Sauvignon (b) are shown. Each bar contains all genes that fall into the same functional class, and induced and repressed genes are denoted by the red and green bars, respectively. Known and putative functions of virus-induced genes were deduced from a TBLASTX search of their DNA sequences against the non-redundant GenBank protein database. High scoring pairs with expected E values less than 10−15 were considered significant. Red and green columns represent induced and repressed genes, respectively

Full size image

The most highly induced genes of virus-infected Carménère and Cabernet-Sauvignon plants are shown in Table 1 and Table 2, respectively. Among these genes, there is an important group that codes for stress- and defense-related proteins. The lipid transfer protein (LTP) that is specifically related to virus response is induced in both cultivars. Stress-responsive proteins like the patatin-like protein, the agenet domain containing protein and MAP kinase phosphatase (MKP1) are also induced. Aging genes like tropinone reductase and harpin-induced family protein (HIN-1), and the detoxifying gene gluthathione S-transferase (GST) were up-regulated. In petunia and maize, diverse GSTs are needed for efficient anthocyanin export from the site of synthesis in the cytoplasm into the vacuole for storage (Mueller et al. 2000). Over-expression of GST may correlate to the fact that GLRaV-3 infected leaves of grapevines usually turn red. Several developmental NAM family NAC2 proteins that are related to biotic and abiotic stress, among other genes related to sugar metabolism like sugar transporters and glycosyl transferase protein, are also induced. It is noteworthy that in Carménère cultivar there is an important induction of sugar, oligopeptides, and hormone transporters (Table 1).

Table 1 Induced genes in Carménère plants during viral infection with GLRaV-3
Full size table
Table 2 Induced genes in Cabernet-Sauvignon plants during viral infection with GLRaV-3
Full size table

Gene induction is mostly associated with cell membrane and endomembrane system, while repressed genes are related to the chloroplast

Viral responsive host genes were classified according to cellular component. A relevant part of the induced genes turned out to be related with the membranous system, especially in the Carménère cultivar (Fig. 4). Genes related to the cell membrane (30.5%) and endomembrane system (23.5%) represent more than half of this group of induced genes. A more detailed analysis revealed that most of them code for proteins involved in the transport function. These include transporters of hormones (auxines and citoquinines), lipids, sugars, and oligopeptides. Also, this category contains genes of proteins associated with the remodeling of the cell wall like extensin and hidroxiproline-rich proteins that are anchored to the cell membrane.

Fig. 4
figure 4

Cellular component involved in grapevine cv. Carménère response to viral infection. The figure shows the percentages of induced (a) and repressed (b) genes during viral infection of Carménère grapevine plants distributed into cell localization categories. Only genes with differential expression of at least twofold were graphed, corresponding to 52 induced genes and 289 repressed genes. To carry out the cellular component assignment, grapevine sequences (TCs from TIGR Grape Gene Index, (http://www.tigr.org/tdb/tgi), matching to probe sets on the array were aligned to Arabidopsis genome (WU-BLAST 2.0) and then, all loci identifiers were used to obtain the GO annotation and functional characterization (http://www.geneontology.org)

Full size image

On the other hand, 59% of the repressed genes are associated with the chloroplast in Carménère cultivar (Fig. 4). An important proportion of these genes codes for proteins associated with metabolic functions. These include photosynthetic proteins as well as photosystems constituents and chlorophyll biosynthesis enzymes. This group contains also ATP-dependent chloroplast protease and chaperones, such as the DNAJ domain-containing protein and the co-chaperone grpE. Some genes associated with the stress-oxidative-detoxification system are also part of this group.

Common response against infection with GLRaV-3 between cultivars

A comparison by means of ANOVA between gene changes in Carménère and Cabernet-Sauvignon plants that occur during viral infection with GLRaV-3 indicated that there were no significant differentially expressed genes. A global view of functions affected by the pathogen showed slight differences between cultivars. However, only 93 in common genes showed significant differences between healthy and virus-infected plants in both cultivars. This group represents the total of statistically significant genes that are responding to the infection without considering the change magnitude. Common genes that were induced and repressed, considering a twofold change cut-off, were functionally classified, and they are listed in Table 3. A great number of genes specifically related with photosynthesis were repressed in both cultivars

Table 3 Overlapping of viral responses of grapevine cultivars Carménère and Cabernet-Sauvignon
Full size table

Verification of microarray data

The over-expression of genes belonging to the biological functions of cell rescue, defense, death, and aging was validated using semi-quantitative RT-PCR. Genes evaluated were: lipid transfer protein (LTP, TC46011), tropinone reductase (Trop, TC40352), and harpin-induced protein 1 (HIN1, TC40876). It was shown that GPDH behaves as a housekeeping gene in a broad repertory of biotic and abiotic conditions (Mahalingam et al. 2003). The selection of this gene as constitutive control was based on its expression level and the fact that no significant differences were found between virus-infected and control grapevine plants in our experimental conditions. To estimate the synthesized product, agarose gel images were quantified and the fold-change relative to a healthy control for each gene was calculated. The induction of LTP, Trop, and HIN 1 was confirmed in these plants by semiquantitative RT-PCR (Fig. 5).

Fig. 5
figure 5

Change expression of defense genes during viral infection. A verification of changes of gene expression of defense related genes in Vitis vinifera Carménère plants by semiquantitative RT-PCR. a Gene-specific primers amplify fragments of 300 bp. All amplifications were done in similar conditions. b Quantification of amplified products was done using Scion Image beta 4.0.2 software (Scion Corporation). Fold-change relative to a healthy control for each gene was calculated with GPDH gene as constitutive control

Full size image

Discussion

Transcript abundance of a broad spectrum of cellular functions is mostly repressed in compatible viral infection in grapevine

The study of a systemic viral infection in the cultivars Cabernet-Sauvignon and Carménère showed that a wide spectrum of grapevine biological functions was affected (Fig. 2). This reflects the fact that viruses are obligate intracellular parasites and that they must establish tight interactions with host cells, causing a general effect in cell physiology.

An overall decrease in transcript abundance was found in response to the compatible viral infection in grapevine and a greater number of repressed genes were obtained in comparison with the up-regulated group (Tables S1 and S2). A host gene shut-off is generally observed upon viral infection in animal and plants (Aranda and Maule 1998; Itaya et al. 2002). This phenomenon, associated with virus replication, is selective over a number of plant genes, while others escape from it or are induced as a result of the infection. It is notorious that the unknown biological processes constitute an important group of affected genes (Figs. 2 and 3). The low similarity of some Vitis gene sequences to Arabidopsis and others known species could explain in part the abundance of genes that belong to this category.

An important number of down-regulated genes represent protein products associated with chloroplast (Fig. 4). Specifically, genes that code for photosystem proteins and enzymes of the biosynthesis of photosynthetic pigments were repressed. A decrease in photosynthetic activity and chlorosis commonly occurs as a symptom of different viral diseases. In relation to this, viral systemic infection causes a depletion of the photosystem II core complex in tobacco leaves that can correlate with the chlorosis response (Lehto et al. 2003).

Grapevine plants activate a defense response against the systemic viral infections

The magnitude of gene expression changes obtained in our grapevine experiments is less dramatic than resistance-associated responses, and is in accordance with previous studies of systemic viral infections in Arabidopsis (Golem and Culver 2003; Ishihara et al. 2004). During an active defense response, biochemical and cell-structural changes destined to restrict the pathogen spread are induced and a systemic acquired response (SAR) mediated by salicylic acid is activated throughout the plant. Pathogenesis-related proteins (PRs), stress-related and detoxifying enzymes genes are straightly associated with this condition. In viral compatible infections, these genes vary along the infection progress (Ishihara et al. 2004) even when some of them are clearly induced (Whitham et al. 2003; Ishihara et al. 2004).

In our case, the defense, cell rescue, and aging functional group contains both up-regulated and repressed genes (Fig. 3). Thus, the host is not passive while systemic infection occurs and some grade of defense is activated, as has been observed in other systemic viral infections (O’Donnell et al. 2003). The LTP gene appears induced in infected tissues of both cultivars (Tables 1, 2 and 3). The mutant dir1 (defective in induced resistance 1) is unable to develop SAR, and interestingly the wild-type DIR1 share sequence similarity to LTPs (Durrant and Dong 2004). Thus, this protein could be involved in the generation of a mobile signal during the SAR. LTPs were associated with hypersensitive response (HR) in pepper and necrotic viral symptoms in Arabidopsis (Sohal et al. 1999; Park et al. 2002), but their role in compatible interaction has not been shown before.

Beta-1, 3-glucan, or callose synthase gene (CALS1) takes part in the response against wounding and in SAR mediated by salicylic acid in Arabidopsis (Ostergaard et al. 2002). Callose synthesis increases during the plant defense against viruses as part of the HR. In compatible infection, callose synthesis increases, but the reaction is not enough to stop the infection (Ehrenfeld et al. 2005). In the grapevine viral infection, the enzyme beta-1, 3-glucanase that hydrolyzes callose, is also induced, as was observed in Arabidopsis in response to several viruses (Whitham et al. 2003) and also in Populus infected with PopMV (Smith et al. 2004).

The repression of some defense genes may be related to the prevalence of GLRaV-3 infection in the grapevine plants over several years. Lactoylglutathione lyase is involved in glutathione-based detoxification of methylglyoxal during plant responses to hormones and either salt or osmotic stresses (Singla-Pareek et al. 2003). Grapevine infected plant showed a lower expression of this gene and probably imply a lower ability to face stressing environmental factors and thus, affect the general condition of the infected plants.

Overlapping between leaf senescence and plant response to virus

It has long been proposed that some overlap exists between leaf-senescence and pathogen-defense programs, but the extent of it remains to be determined (Quirino et al. 2000). Some Carménère virus-induced genes support this overlap concept, and examples include tropinone reductase, HIN1, aspartyl protease, and the above-mentioned LTPs. Tropinone reductase encodes a short-chain alcohol dehydrogenase and has been described as the senescence-associated gene 13 (Weaver et al. 1998). The biological function of HIN1 is unknown but shares sequence identity with NDR1 protein, which is required for the HR in Arabidopsis against bacteria, fungi, and viruses, probably associated with cell death (Pontier et al. 1999). Aspartyl proteases have been related with protein processing at different stages of plant development, including leaf senescence (Quirino et al. 2000; Simoes and Faro 2004). Besides its role in the recycling of nutrients during senescence, aspartyl proteases are related with the degradation of PR proteins in tomato and tobacco, probably preventing an over-accumulation of them. Aspartyl proteases may also play an important role during chloroplast degradation, explaining the chlorosis symptoms and the lower photosynthetic activity observed in some virus-infected plants.

Transport is a function highly affected by viral systemic infection in grapevine

Genes associated with the membrane and endomembrane systems are expressed in both cultivars after the infection with GLRaV-3. This effect is clearly observed in Carménère cultivar (Fig. 4). Many of these genes are associated with the transport function. For example, the expression of monosaccharides, and other sugars transporters are induced. A clear induction of genes that code for sugar transporters has been reported in Arabidopsis and tomato in response to fungal infections (Fotopoulos et al. 2003; Garcia-Rodriguez et al. 2005). In the last case, a re-direction of sugars to regions colonized by the pathogen has been suggested. An alteration in source–sink balance could explain the delay in berries ripening and anthocyanins accumulation that accompany several viral infections as a relation between these two processes has been described in grapes (Cakir et al. 2003). Additionally, it has been reported that virus-infected grapevine plants produce juice must with an important decrease in sugars and its organoleptic properties (Borgo and Angelini 2002).

A group of the induced genes corresponds to proton-dependent oligopeptide transporters (POT) (Table 1) that constitutes a family of proteins associated with membranes (Lubkowitz 2006). As POTs are involved in auxin transport, an up-regulation of this kind of protein alters the homeostasis of this hormone, which has been implicated with systemic viral diseases. For example, TMV replicase interacts with PAP1 protein, a putative regulator of auxin response, interfering with the hormone activity in Arabidopsis and probably participating in the induction of specific disease symptoms (Padmanabhan et al. 2005). Furthermore, it has been suggested that the plant defense response during infection with viruses could be regulated through auxin homeostasis (Mayda et al. 2000), and viruses can alter the steady state of this hormone, leading to morphogenetic alterations, such as the leaf distortion characteristic of leaf roll disease (Pennazio and Roggero 1996). In agreement with this, the up-regulation of an auxin efflux carrier protein was observed in virus-infected Carménère plants.

Host genes involved in viral cycle

A successful viral infection requires the completion of events, such as viral genome replication, cell-to-cell movement, and long-distance transport. Some host genes have been implicated in these processes, and, in this sense, the repression of two DEAD/DEAH box RNA helicase and DNAJ heat shock protein is intriguing. It has been demonstrated that the DEAD-box RNA helicase is required for Brome Mosaic Virus (BMV) RNA synthesis (Maule et al. 2002). DNAJ heat shock protein acts as a chaperone and it has been shown to participate in the negative-strand RNA synthesis of BMV (Tomita et al. 2003). Replication mechanism of GLRaV-3, as well as the other members of the Closteroviridae family, involves the generation of multiple subgenomic RNAs through the negative-strand RNA synthesis (Martelli et al. 2002). Thus, a plant’s response to viral infection could occur by a lower expression of DEAD/DEAH box RNA helicase and DNAJ heat shock protein, to attempt to counteract viral replication.

Another important aspect related to infection is viral movement. It was demonstrated that pectin methylesterase (PME) is necessary for cell-to-cell and systemic movement of TMV in tobacco plants (Chen and Citovsky 2003; Chen et al. 2000). A PME gene is induced in infected Carménère plants. PME role in viral spread of TMV could be derived from its direct interaction with the tobamovirus movement protein. Although none of the ORFs described in GRLaV-3 genome has been assigned as a movement protein, it is possible that some ORFs with unknown functions may act in viral spread (Ling et al. 2004).

In GLRaV-3-infected grapevine plants we observed a repression of the expression of the heat shock protein 70, Hsp70 gene, which contrasts with its induction in other viral infections (Maule et al. 2002; Whitham et al. 2003; Aparicio et al. 2005). Members of the Closteroviridae family encode a Hsp70 homologue (Hsp70h) that is necessary for virion assembly and for cell-to-cell movement of a closterovirus (Alzhanova et al. 2001). Hsp70h, located at plasmodesmata, could, therefore, be participating in virion translocation and acting as a chaperone or supplying energy through the Hsp70h-mediated ATP hydrolysis.