Nonhost versus host resistance to the grapevine downy mildew, Plasmopara viticola, studied at the tissue level.

Following inoculation of host and nonhost plants with Plasmopara viticola, the grapevine downy mildew, a histological survey was undertaken to identify the stage where its development is contained in nonhosts and in resistant host plants. Three herbaceous nonhost species, Beta vulgaris, Lactuca sativa, and Capsicum annuum, and three grapevine species displaying different level of resistance (Vitis vinifera [susceptible], Vitis riparia [partially resistant] and Muscadinia rotundifolia [totally resistant]) where inoculated by P. viticola using a controlled leaf disk inoculation bioassay. During the early steps of infection, defined as encystment of zoospores on stomata, penetration of the germ tube, and production of the vesicle with the primary hypha, there was no evidence of a clear-cut preference to grapevine tissues that could attest to host specificity. The main difference between host grapevine species and nonhosts was observed during the haustorium formation stage. In nonhost tissues, the infection was stopped by cell wall-associated defense responses before any mature haustorium could appear. Defense responses in resistant grapevines were triggered when haustoria were fully visible and corresponded to hypersensitive responses. These observations illustrate that, for P. viticola, haustorium formation is not only a key stage for the establishment of biotrophy but also for the host specificity and the recognition by grapevine resistance factors.

Development of a transient expression system in grapevine via agro-infiltration.

Transient expression of genes using Agrobacterium is a powerful tool for the analysis of gene function in plants. We have developed this method for the analysis of genes involved in disease resistance in grapevine leaves. Our research showed that the quality of the plant material, the plant genotype used for agro-infiltration and the presence of additional virulence factors (carried on plasmid pCH32) in the Agrobacterium strain are all important factors for success of the procedure. After optimising these factors, we consistently achieve sufficient acceptable levels of expression of the markers beta-glucuronidase (GUS) and green fluorescent protein (GFP) using vacuum infiltration of grapevine leaves from plants grown in vitro. We used this procedure to investigate the proposed role of stilbenes in defense against grapevine downy mildew (Plasmopara viticola) by transiently overexpressing stilbene synthase in grapevine leaves, before infection with P. viticola. We found that agro-infiltration itself induces the synthesis of stilbenes in grapevine leaves, thus preventing us to test the effect of the overexpression of stilbene synthase in defense. However, our results revealed that agro-infiltration before P. viticola inoculation had an effect on the development of the infection. Further research is required to show whether stilbenes or some other factor are the causal agent restricting pathogen development. The method described here provides and excellent tool to exploit at the many grapevine genomic resources now available, and will contribute to a better understanding of many areas of grapevine biology.

Multiple origins of cultivated grapevine (Vitis vinifera L. ssp. sativa) based on chloroplast DNA polymorphisms.

The domestication of the Eurasian grape (Vitis vinifera ssp. sativa) from its wild ancestor (Vitis vinifera ssp. sylvestris) has long been claimed to have occurred in Transcaucasia where its greatest genetic diversity is found and where very early archaeological evidence, including grape pips and artefacts of a 'wine culture', have been excavated. Whether from Transcaucasia or the nearby Taurus or Zagros Mountains, it is hypothesized that this wine culture spread southwards and eventually westwards around the Mediterranean basin, together with the transplantation of cultivated grape cuttings. However, the existence of morphological differentiation between cultivars from eastern and western ends of the modern distribution of the Eurasian grape suggests the existence of different genetic contribution from local sylvestris populations or multilocal selection and domestication of sylvestris genotypes. To tackle this issue, we analysed chlorotype variation and distribution in 1201 samples of sylvestris and sativa genotypes from the whole area of the species' distribution and studied their genetic relationships. The results suggest the existence of at least two important origins for the cultivated germplasm, one in the Near East and another in the western Mediterranean region, the latter of which gave rise to many of the current Western European cultivars. Indeed, over 70% of the Iberian Peninsula cultivars display chlorotypes that are only compatible with their having derived from western sylvestris populations.

An integrated SSR map of grapevine based on five mapping populations.

A grapevine (mainly Vitis vinifera L., 2n = 38) composite genetic map was constructed with CarthaGene using segregation data from five full-sib populations of 46, 95, 114, 139 and 153 individuals, to determine the relative position of a large set of molecular markers. This consensus map comprised 515 loci (502 SSRs and 13 other type PCR-based markers), amplified using 439 primer pairs (426 SSRs and 13 others) with 50.1% common markers shared by at least two crosses. Out of all loci, 257, 85, 74, 69 and 30 were mapped in 1, 2, 3, 4 and 5 individual mapping populations, respectively. Marker order was generally well conserved between maps of individual populations, with only a few significant differences in the recombination rate of marker pairs between two or more populations. The total length of the integrated map was 1,647 cM Kosambi covering 19 linkage groups, with a mean distance between neighbour loci of 3.3 cM. A framework-integrated map was also built, with marker order supported by a LOD of 2.0. It included 257 loci spanning 1,485 cM Kosambi with a mean inter-locus distance of 6.2 cM over 19 linkage groups. These integrated maps are the most comprehensive SSR-based maps available so far in grapevine and will serve either for choosing markers evenly scattered over the whole genome or for selecting markers that cover particular regions of interest. The framework map is also a useful starting point for the integration of the V. vinifera physical and genetic maps.

Diversification within grapevine cultivars goes through chimeric states.

Vitis vinifera 'Pinot' clones were analysed at 50 microsatellite loci to assess intravarietal genetic diversity. When analysing leaf tissue DNAs, polymorphism mainly resulted from the appearance of a third allele when two were expected for heterozygous loci in a diploid species. The sequencing of the three microsatellite alleles at two loci has confirmed their simultaneous presence in the leaf tissues. A hypothesis explaining the triallelic profiles at a locus is the presence of a periclinal chimera meristem structure, in which genetically different cell layers coexist. The periclinal chimeric state of two Vitis vinifera 'Pinot gris' clones was confirmed by splitting and analysing the genotypes resulting from L1 and L2 cell layers in progeny derived from self-fertilization, in root tissues, and in plants regenerated from somatic embryogenesis. Prevalence of chimerism in polymorphic clones observed in a collection of 145 accessions belonging to 'Pinot gris', 'Pinot noir', Pinot blanc', 'Pinot meunier', and 'Pinot moure' cultivars was demonstrated. The accumulation of somatic mutations and cell layer rearrangements allowed us to deduce the relationships between the various genotypes and to open a way for understanding the diversification process and the phylogeny in the 'Pinot' group.

Mapping 245 SSR markers on the Vitis vinifera genome: a tool for grape genetics.

The aim of the present work was to develop a microsatellite marker-based map of the Vitis vinifera genome (n=19), useful for genetic studies in this perennial heterozygous species, as SSR markers are highly transferable co-dominant markers. A total of 346 primer pairs were tested on the two parents (Syrah and Grenache) of a full sib population of 96 individuals (S x G population), successfully amplifying 310 markers. Of these, 88.4% markers were heterozygous for at least one of the two parents. A total of 292 primer pairs were then tested on Riesling, the parent of the RS1 population derived from selfing (96 individuals), successfully amplifying 299 markers among which 207 (62.9%) were heterozygous. Only 6.7% of the markers were homozygous in all three genotypes, stressing the interest of such markers in grape genetics. Four maps were constructed based on the segregation of 245 SSR markers in the two populations. The Syrah map was constructed from the segregations of 177 markers that could be ordered into 19 linkage groups (total length 1,172.2 cM). The Grenache map was constructed with the segregations of 178 markers that could be ordered into 18 linkage groups (total length 1,360.6 cM). The consensus S x G map was constructed with the segregations of 220 markers that were ordered into 19 linkage groups (total length 1,406.1 cM). One hundred and eleven markers were scored on the RS1 population, among them 27 that were not mapped using the S x G map. Out of these 111 markers, 110 allowed to us to construct a map of a total length of 1,191.7 cM. Using these four maps, the genome length of V. vinifera was estimated to be around 2,200 cM. The present work allowed us to map 123 new SSR markers on the V. vinifera genome that had not been ordered in a previous SSR-based map (Riaz et al. 2004), representing an average of 6.5 new markers per linkage group. Any new SSR marker mapped is of great potential usefulness for many applications such as the transfer of well-scattered markers to other maps for QTL detection, the use of markers in specific regions for the fine mapping of genes/QTL, or for the choice of markers for MAS.

Complete sequence of Tvv1, a family of Ty 1 copia-like retrotransposons of Vitis vinifera L., reconstituted by chromosome walking.

A chromosome-walking strategy was used to sequence and characterize retrotransposons in the grapevine genome. The reconstitution of a family of retroelements, named Tvv1, was achieved by six successive steps. These elements share a single, highly conserved open reading frame 4,153 nucleotides-long, putatively encoding the gag, pro, int, rt and rh proteins. Comparison of the Tvv1 open reading frame coding potential with those of drosophila copia and tobacco Tnt1, revealed that Tvv1 is closely related to Ty 1 copia-like retrotransposons. A highly variable untranslated leader region, upstream of the open reading frame, allowed us to differentiate Tvv1 variants, which represent a family of at least 28 copies, in varying sizes. This internal region is flanked by two long terminal repeats in direct orientation, sized between 149 and 157 bp. Among elements theoretically sized from 4,970 to 5,550 bp, we describe the full-length sequence of a reference element Tvv1-1, 5,343 nucleotides-long. The full-length sequence of Tvv1-1 compared to pea PDR1 shows a 53.3% identity. In addition, both elements contain long terminal repeats of nearly the same size in which the U5 region could be entirely absent. Therefore, we assume that Tvv1 and PDR1 could constitute a particular class of short LTRs retroelements.

Comparison of enzyme-linked immunosorbent assay (ELISA) with dot hybridization using 32P- or 2-acetylaminofluorene (AAF)-labelled cDNA probes for the detection and characterization of beet necrotic yellow vein virus.

Beet Necrotic Yellow Vein Virus (BNYVV) was detected by enzyme-linked immunosorbent assay (ELISA) and RNA/DNA dot hybridization using either radiolabelled or non-radioactive probes. Dot hybridization specifically distinguished isolates that could not be distinguished by ELISA. The detection thresholds for ELISA, hybridization with non-radioactive probes and hybridization with radiolabelled probes were 2 ng, 0.2 ng, 0.02 ng of purified virus, respectively. Dot hybridization with non-radioactive probes could be performed on crude infected beet root extracts, thus providing a useful tool for monitoring BNYVV infection and for routine testing in plant breeding programs.

Effect of beet necrotic yellow vein virus RNA composition on transmission by Polymyxa betae.

Beet necrotic yellow vein virus (BNYVV) is naturally transmitted by the soil-borne fungus Polymyxa betae and usually remains confined to the roots of infected sugarbeets. In naturally infected sugarbeets the virion RNA always consists of four components which are uniform in size in different isolates but when BNYVV is propagated by mechanical inoculation to leaves of Chenopodium quinoa the two smallest RNA components, RNA-3 and -4, may undergo deletion or disappear from the isolate, suggesting that they are only essential for the natural mode of infection. To test this hypothesis, several C. quinoa isolates of BNYVV with different RNA-3 and -4 contents have been retransmitted to sugarbeet root via P. betae. The results show that the two isolates containing no detectable full-length RNA-3 and -4 are poorly transmitted and that cases of successful infection are associated with the reappearance of full-length RNA-3 and -4.