Experimental

Genotyping analyses allowed us to estimate that there are approximately 5000 grape varieties (This et al., Reference This, Lacombe and Thomas2006). The cultivar is the basic element in grape growing and wine production, as well as very important from a legal perspective.

Today, microsatellite molecular markers (simple sequence repeats (SSRs)) have become a complementary tool to traditional ampelography for cultivar identification, and techniques capable of speeding up the SSR genotyping are highly desirable.

High-quality genomic DNA is required for most molecular analyses. Grapevine is rich in secondary metabolites, and optimal quantity and quality DNA is usually obtained from young leaves. However, leaves are not always available, and thus it becomes convenient to analyse the DNA from other plant sources, such as wood or grape. In the latter case, it is more difficult to obtain sufficiently pure DNA due to a higher concentration of DNA polymerase inhibitors such as polyphenols and polysaccharides. Nevertheless, new engineered DNA polymerases have recently been developed, with improved tolerance to plant PCR inhibitors and ability to amplify the DNA directly from crude samples, bypassing the purification steps.

Moreover, well-balanced multiplex PCR allows the simultaneous amplification of more SSR loci per reaction, with a considerable saving of time and effort; however, the optimization of the system is crucial and the range size of the alleles and the dynamics of the primers have to be considered (Ibáñez et al., Reference Ibáñez, Vargas, Palancar, Borrego and De Andrés2009).

In this context, we developed an innovative approach for grapevine genotyping by combining a DNA polymerase able to tolerate PCR inhibitors with one multiplex PCR at 11 SSR loci, applied directly on crude tissues. This method was tested on a range of different plant sources such as leaf, root, wood, fresh and frozen berry, stalk and must. Four wine grape cultivars and four rootstocks from the CRA-VIT repository were used as a sample source; in addition, bunches of four table commercial cultivars were bought at a local supermarket. These samples were chosen to represent a large variability in the chemical composition of different tissues and varieties (Table 1).

Table 1 Grapevine varieties and tissues used for molecular analysis

The template, approximately 0.3 mm diameter of crude sample or 1 μl of must, was placed directly into a 0.2 ml Eppendorf tube containing 50 μl of PCR mix using the KAPA3G Plant PCR Kit (Kapa Biosystems, Inc., Boston, MA, USA), and amplified as detailed in the Supplementary Material (available online only at http://journals.cambridge.org).

The routine grapevine genotyping at the CRA-VIT uses 11 SSRs, six proposed by This et al. (Reference This, Jung, Boccacci, Borrego, Botta, Costantini, Crespan, Dangl, Eisenheld, Ferreira-Monteiro, Grando, Ibáñez, Lacombe, Laucou, Magalhães, Meredith, Milani, Peterlunger, Regner, Zulini and Maul2004) and adopted by OIV (2009) in the descriptor list for grape varieties and Vitis species, and the remaining five selected to enhance the resolving power (Bowers et al., Reference Bowers, Dangl and Meredith1999; Crespan, Reference Crespan2003; Welter et al., Reference Welter, Göktürk-Baydar, Akkurt, Maul, Eibach, Töpfer and Zyprian2007). Therefore, the multiplex PCR was set up to amplify these 11 SSR loci in a single reaction. Forward primers were labelled with 6-FAM, NED, PET or VIC fluorochromes (Supplementary Table S1, available online only at http://journals.cambridge.org) and balanced with the criteria reported in the Supplementary Material (available online only at http://journals.cambridge.org). Amplification products were separated on an ABI Prism 3110xl genetic analyser (Life Tech, Foster City, CA, USA). After data collection, genotyping analysis was performed with GeneMapper software version 3.0 (Life Tech), giving rise to four separate panels of signals (Fig. 1). Genotyping data are reported in Supplementary Table S2 (available online only at http://journals.cambridge.org).

Fig. 1 Example of the multiplex PCR molecular profile of 11 SSR loci obtained with the proposed methodology. (a) NED-, (b) 6-FAM-, (c) PET- and (d) VIC-labelled SSR loci. Grey and red stripes represent the home-made bin set produced with reference varieties.

The experimental results confirm that this method works very well, amplifying the entire set of 11 SSRs in a single step directly from all plant tissues tested. In addition, PCR specificity and genotyping accuracy are comparable with those obtained with conventional methods that require DNA purification steps. It is important to mention that the ratio between the amount of crude sample and PCR mix volume is crucial, because no amplification is obtained by increasing the raw material in the reaction mix.

To perform the varietal identification, the SSR profile was compared with the CRA-VIT molecular database (unpublished), which currently contains 1440 unique genotype records. The probability of identity computed with Cervus software version 3.0 (Kalinowski et al., Reference Kalinowski, Taper and Marshall2007) using the 11 SSR markers gave a very close-to-zero outcome (i.e. 3.71 × 10^− 16). It is important to mention that SSR markers are not useful to distinguish varieties derived from somatic mutations (Sefc et al., Reference Sefc, Lefort, Grando, Scott, Steinkellner, Thomas and Roubelakis-Angelakis2001). After the comparison, all known samples of SSR profiles matched those of the corresponding variety present in the CRA-VIT database. About the unknown table grapes, samples 1, 3 and 4 were identified as Sugraone, Red Globe and Regina, respectively, whereas sample 2 remained unknown. This last variety could be identified in the future, based on present genotyping results and new experimental or literature data.