Simple sequence repeat (SSR) vs. sequence-related amplified polymorphism (SRAP) markers for Cynara cardunculus characterization

A little is known about the genetic variability present in globe artichoke, cultivated and wild cardoons. This knowledge is very important for efficient genetic resources utilization, and to gain a better understanding of genetic structure of this botanical varieties. With the aims to determine genetic distances between Cynara cardunculus accessions and to compare two molecular markers systems for their efficiency to differ between botanical varieties, a molecular characterization of sixteen accessions from different geographical origins was performed. Seven SSR and seven SRAP markers were used for varieties characterization and to calculate genetic distances between them. Both distance matrices were subjected to cluster analysis. Exclusive SSR alleles were found for globe artichoke and for wild cardoon, but non exclusive alleles were found for cultivated cardoon. For both markers systems two major groups were identified, one of them included mostly globe artichoke accessions and the other one grouped mainly cardoons. The differences observed in the sub-cluster conformation with each marker systems may be due to intrinsic characteristics of the markers. Concluding, both kind of molecular markers are valuable tools for studying genetic distances between C. cardunculus accessions although they give different information. Nevertheless, SSR electrophoretic profiles are simpler to score than SRAP markers because they consist of just a few bands. As well, bands are highly informative because of the great number of alleles existing in population and they are codominant markers. In addition, SSR’s use would reduce time and costs.

Globe artichoke heads or capitulum, which are immature composite inflorescences, are the edible parts of the plant and are used as a fresh, frozen or canned delicacy all over the world. Its cultivation is largely diffused in European Mediterranean countries and, to a lesser extent in South America, United States and, more recently, also in China (Bianco, 2005). The edible parts of cultivated cardoon are fleshy stems. It is cultivated in Spain, Italy and the south of France, where it is used in traditional dishes (Lanteri and Portis, 2007). In Argentina it is cultivated and consumed by European immigrant's descendants.
Alternative uses of these crops have been described in the last years, artichoke leaf extracts have been reported to inhibit cholesterol biosynthesis and to prevent cardiovascular disease. Also, artichoke and cardoon achenes are oil sources, and their biomass could be used for forage, paper pulp or biofuel (Maccarone et al., 1999;Gominho et al., 2000;Curt et al., 2002;Quilho et al., 2004).
In contrast to other horticultural crops, no intensive breeding programs have been carried out for globe artichoke or cultivated cardoon (Sonnante et al., 2008). For this reason, a better knowledge of C. cardunculus genetics and variability is crucial in order to plan new breeding strategies, for efficient germplasm resources utilization and to maintain genetic diversity, which sustains long-term selection responses and reduces vulnerability (Sorrels and Wilson, 1997;Troyer et al., 1998;Jana, 1999).
Several authors analyzed the variability among C. cardunculus accessions using morphologic traits and bio-chemical markers Porceddu et al., 1976;Miccolis et al., 1989;Hammouda et al., 1993;Asprelli et al., 2001). However, some of these markers are highly influenced by environmental conditions and give limited information, thus it is essential to include molecular characterization in diversity studies.
There are several kinds of molecular markers, each one brings different information and have its own advantages and disadvantages. Microsatellites, also known as simple sequence repeat (SSR) are tandem repeats of 1 to 6 bp which occur frequently in most eukaryotes genomes and can be highly informative since they are multiallelic and greatly reproducible. To amplify a particular SSR it is necessary to know the sequence flanking it to design usefull SSR primers. This can be laborious, time consuming and costly, then, the use of SSR technology depends usually, on the availability of suitable primers. Several SSR markers have been developed for C. cardunculus (Acquadro et al., 2003(Acquadro et al., , 2005a(Acquadro et al., ,b, 2009), so it is not a limiting variable in this species.
On the other hand, sequence-related amplified polymorphism (SRAP) target open reading frames (ORFs). Primers consist of core sequences, which are 13 to 14 bases long, where the first 10 or 11 bases starting at the 5' end, with no specific constitution, followed by the CCGG sequence in the forward primer (to hybridize with exons which are CG rich regions) and AATT sequence in the reverse primer which target introns (AT rich). These core sequences are followed by three selective nucleotides at the 3' end which are variable from a primer to another (Li and Quiros, 2001). The advantages of this technology are that it is not necessary to know any sequence of genome under study and that this kind of markers is multilocus.
The aims of this study were to determine genetic distances between sixteen C. cardunculus accessions and to compare SSR with SRAP marker systems for their effectiveness in differing between globe artichoke and wild and cultivated cardoon.

Material and methods
Sixteen accessions of Cynara cardunculus L. including globe artichoke, wild and cultivated cardoon (Table 1), were sown in a completely randomized design in the Experimental Station of Rosario National University, located at Zavalla (33°1' S; 60°53' W), Argentine. Since this study did not aim to determine the presence of genetic variability within the accessions, five plants bulk of each accession were analyzed in order to represent the putative genetic variability present within accessions. Young fresh leaves of adult plants were ground to powder in liquid nitrogen and then genomic DNA was extracted using the PureLink Plant Total DNA Purification Kit (Invitrogen, Carlsbad, CA, USA). The resulting DNA was resuspended in TE buffer (10 mM Tris-HCl, 1 mM EDTA, pH 8.0). Then it was quantified with commercial kit «QubitTM Flourometer (Invitrogen)» and stored at -20°C.

SRAP protocols
Seven primers combination (Me1/Em4, Me1/Em5, Me2/Em4, Me3/Em2, Me3/Em5, Me4/Em3, Me4/Em4) were tested (Li and Quiros, 2001). Each PCR reaction sample was performed with 20 ng DNA, 5 mM dNTPs, 50 mM MgCl 2 , 10 µM of each primer, 5.2 µL 10X Taq buffer and 4 U Taq polymerase (PB-L, Quilmes National University) in a final volume of 20 µL. Samples were amplified by PCR with the following thermal profile: 5 min at 94°C, 3 cycles of 1 min at 94°C, 1 min at 35°C and 1 min at 72°C, a 35 cycles of 1 min at 94°C, 1 min at 50°C and 1 min at 72°C, with a final step of 1 min at 72°C. The amplification products were also visualized in denaturing polyacrylamide gels with silver staining. Bands were scored as present (1) or absent (0). For each marker system, genetic distances were calculated according to standardized Jaccard's distance index (JDI) (Di Renzo et al., 2001). Both JDI matrices were subjected to cluster analysis and a dendrogram was created for each one using, in both cases, the Ward's method for InfoStat statistical software (Di Renzo et al., 2001). To test the dendrogram goodness of fit, the cophenetic correlation between the similarity matrix and the corresponding cophenetic matrix (computed from the tree matrix that generate the dendrogram) was calculated (Rohlf, 1974).

Results and discussion
Ten microsatellites were initially tested; three of them did not show robust amplification profiles and were discarded. Therefore, the analysis of the sixteen Cynara cardunculus accessions was performed with seven microsatellites, which gave one or two alleles per accession. The whole primers tested detect 25 alleles across the 16 accessions, ranging from 1 (CMAL-24) to 5 (CMAL-21 and CMAL-06) alleles per locus, with a 92% of polymorphism. Microsatellites are regions with high mutation rate, so such high percentages of polymorphism are very frequent (Matus and Hayes, 2002). Acquadro et al. (2005b) observed 65.9% of polymorphic alleles assessing genetic distances in a C. cardunculus population with SSR markers.
We found two SSR alleles exclusive for globe artichoke. One of them is a fragment of 140 bp obtained with the primers CMAL-21 and the other one is a fragment of 310 bp obtained with CMAFLP-04. Another two exclusive alleles were found for wild cardoon with the primers CMAL-21 and CMAFLP-07 (150 and 200 bp, respectively). No exclusive alleles were detected for cultivated cardoon. This fact may be a consequence of loose of genetic identity as a result of reiterating intercrossing between cultivated cardoon and the others botanical varieties (wild cardoon or globe artichoke). Acquadro et al. (2005b) by SSR markers amplified 6 distinctive alleles in globe artichoke and 26 in wild cardoon, but neither exclusive alleles were identified in cultivated cardoon, since each allele observed in cultivated cardoon was shared by at least one globe artichoke accession.
JDI obtained from SSR data ranged from 0.35 to 0.96 (Table 3). The lowest values were observed between both globe artichoke Brazilian accessions (Feltrin Roxa and Feltrin Verde) and between two Argentinian globe artichoke accessions (Gauchito and 51A 3 , both from the local breeding program). The highest JDI values were found between a wild cardoon accessions (R9) and two Argentinian globe artichoke varieties (Ñato and Gauchito). The average JDI was 0.73, indicating that the germplasm under study is fairly rich in diversity (Chen and Du, 2006).
Cluster analysis was able to identify two major groups. The cophenetic correlation was higher than 0.70 which indicates an acceptable goodness of fit of the similarity indices. One of the obtained groups (CI) included mostly globe artichoke accessions, whereas, the other one (CII) grouped mainly cardoons, with some exceptions (Fig. 1).
Cluster 1 (CI) was sub-divided into three sub-clusters. Each one was performed with accessions with common geographical origin, thus Feltrin Roxa and Feltrin Verde (from Brazil) were grouped together with the lowest JDI (0.35); Violetto Precocce and Romanesco, both Italian varieties, perform another sub-cluster (JDI = 0.63) together with a cultivated cardoon Semence (Argentine). Finally, Esmeralda, 51A 3 , Gauchito and Ñato, which are all Argentinean varieties obtained from the The analysis performed with SRAP data (270 polymorphic bands) showed an average JDI = 0.74 (Table 3) similar to that obtained with SSR data. Nevertheless, these JDI values presented a minor range of dispersion as it has been expected with data coming from coding sequences which tend to be more conserved. JDI's ranged between 0.58 (between cultivated cardoon Florensa and wild cardoon R9) and 0.85 (between Gauchito and F3 or Gauchito and R9). These results agree with those obtained with SSR in which Gauchito and R9 also show the largest distances from one to the other.
The dendrogram obtained with these JDI data also contained two main clusters (Fig. 2) in which cardoons and globe artichoke accessions appear separated one from the others (cophenetic correlation coefficient = 0.78). In this sense, Cluster 1 (CI) included all the cardoons accessions together with Feltrin Verde; whereas Cluster 2 (CII) included only globe artichokes, most of them grouped in an only one sub-cluster, except Ñato and Violetto Precocce which showed a JDI = 0.76 and conformed another sub-cluster.
Accessions F1, F2 and F3 together with Feltrin Verde perform a sub-cluster into the cluster 1, whereas the cultivate cardoons (Semence and Florence) and the wild cardoon accession grouped in another sub-cluster. Semence and Florence are both commercial varieties, whereas F1, F2 and F3 are three accessions collected in local producers' fields.
Both markers systems showed clusters with two main groups, one mostly conformed to globe artichoke accessions and the other one with cultivated cardoons, SSR vs. SRAP markers for Cynara cardunculus characterization 457 Table 3. Jaccard's distances index matrix between the 16 accessions obtained by SRAP (above the diagonal) and SSR (below the diagonal)  nevertheless dendrograms were different in branching within each group. The single wild cardoon included in the assay (R9) grouped with cultivated cardoons in both dendrograms. DNA studies (Lanteri et al., 2004;Acquadro et al., 2005a), cytogenetic-and isozyme-based analysis  have all confirmed that wild cardoon is the ancestor of both cultivated C. cardunculus forms, which evolved independently under the influence of different anthropogenic selection criteria: globe artichoke for its capitula and cardoons for its succulent leaves. On the other hand, domestication of globe artichoke took place in the beginning of the first millennium while domestication of cardoon took place in the first half of the second millennium (Sonnante et al., 2007). Due to the low time since from both botanical varieties separation, it is reasonable that cultivated cardoons share more alleles with the wild progenitor than globe artichoke. In addition, both cardoon forms are full cross-compatible and seed propagated what may contribute to maintain a rather similar genetic background.
The differences observed in the sub-cluster conformation with each marker systems may be due to intrinsic characteristics of the markers. As Li and Quiros (2001) have pointed out, SRAP target ORFs, sequences which tend to be more conserved between genotypes than microsatellites which are generally non-coding regions, then, they are more variable. Clearly, the information given by both marker systems is different, which, results in not coincident dendrograms.
It is probable that if the number of SSR markers increase, the three botanical varieties identification will be improved. Acquadro et al. (2005b) genotyping 36 C. cardunculus accessions with 14 SSR, obtained three clusters, one of them contained only globe artichokes, other conformed to wild cardoons and the third one contained exclusively cultivated cardoon accessions.
SRAP technology give more bands per primer combination tested than SSR because of the random design of SRAP primers; possibly because of it, when we tried to obtain a dendrogram using the whole data acquired with both marker systems the tree was similar to the SRAP one.
Concluding, both kind of molecular markers are valuable tools for study genetic distances between globe artichokes and cardoons although they give different information. Nevertheless, SSR electrophoretic profiles are simpler to score than that resultant of SRAP markers because they consist of just a few bands. As well, bands are highly informative because of the great number of alleles existing in population and they are codominant markers. Also, SSR brings the possibility to perform multiplex-PCR increasing the information obtained in each experiment, reducing time and costs.