Transferability and polymorphism of barley EST-SSR markers used for phylogenetic analysis in Hordeum chilense

Background Hordeum chilense, a native South American diploid wild barley, is a potential source of useful genes for cereal breeding. The use of this wild species to increase genetic variation in cereals will be greatly facilitated by marker-assisted selection. Different economically feasible approaches have been undertaken for this wild species with limited direct agricultural use in a search for suitable and cost-effective markers. The availability of Expressed Sequence Tags (EST) derived microsatellites or simple sequence repeat (SSR) markers, commonly called as EST-SSRs, for barley (Hordeum vulgare) represents a promising source to increase the number of genetic markers available for the H. chilense genome. Results All of the 82 barley EST-derived SSR primer pairs tested for transferability to H. chilense amplified products of correct size from this species. Of these 82 barley EST-SSRs, 21 (26%) showed polymorphism among H. chilense lines. Identified polymorphic markers were used to test the transferability and polymorphism in other Poaceae family species with the aim of establishing H. chilense phylogenetic relationships. Triticum aestivum-H. chilense addition lines allowed us to determine the chromosomal localizations of EST-SSR markers and confirm conservation of the linkage group. Conclusion From the present study a set of 21 polymorphic EST-SSR markers have been identified to be useful for diversity analysis of H. chilense, related wild barleys like H. murinum, and for wheat marker-assisted introgression breeding. Across-genera transferability of the barley EST-SSR markers has allowed phylogenetic inference within the Triticeae complex.


Background
Wild species constitute a potential source of genetic variation for cultivated species. Besides, they can be analyzed to answer the long-lasting questions concerning the origins, evolution and spread of major agricultural crops of the world. Recently, there has been considerable progress in plant genomics, leading to novel molecular breeding tools to reduce the costs and to simplify the assays. Plant genome research has been focused on the major crops and model species and a vast amount of genomic information has been accumulated. This information will provide an opportunity to use it as sources of information for thousands of minor grass species [1].
Hordeum chilense Roem et Schultes is a native South American diploid perennial wild barley (2n = 2x = 14), included in the section Anisolepsis [2]. It belongs to a heterogeneous group of South American Hordeum species and it is one of the species of the genus Hordeum with a high potential for cereal breeding purposes, given its high crossability with other members of the Triticeae tribe. H. chilense was used to obtain fertile amphiploids with wheat of different ploidy levels (diploid, tetraploid and hexaploid). These amphiploids were named Tritordeum and are the basic genetic material for using H. chilense genetic variability in wheat breeding [3]. H. chilense has agronomically interesting characteristics, like high carotenoid content, biotic and abiotic stress resistance, and variability for seed storage proteins [4]. A new cytoplasmic male sterility (CMS) source in bread wheat using the cytoplasm of H. chilense has also been reported [5].
The characterization of genetic variability in wild species like H. chilense and the development of tools to introduce it into cultivated crops are important plant breeding goals. The analysis of DNA sequence variation is of major importance in genetic studies. In this context, molecular markers are useful tools for assaying genetic variation, and have greatly enhanced the genetic analysis of crop plants. High development costs make it impractical to develop molecular markers directly from wild species like H. chilense. However, if markers developed in related crop species can be used, genetic analysis of H. chilense could be advanced rapidly [6]. Across-species transferability of SSRs derived from EST databases is greater than that of SSRs derived from enriched genomic DNA libraries, as they originate from expressed regions and therefore they are more conserved across a number of related species than non-coding regions [19]. They have shown to be useful for comparative mapping across species, comparative genomics, and evolutionary studies and they have been shown to posses a higher potential for inter-specific transferability than genomic SSRs [15,16,[20][21][22][23]. On the other hand, they are expected to be less polymorphic within the species due to its conserved nature [19]. In summary, EST-SSR has provided a valuable source of new PCR-based molecular markers in cereal crops.
We have tested the transferability of 82 EST-SSR markers developed in barley [24,25], and their potential use as new molecular tools for introgression, variability and phylogenetic analysis of the H. chilense genome. The chromosomal locations of the transferred and polymorphic EST-SSRs were assigned using wheat chromosome addition lines [26].
The grass family (Poaceae) is formed by 600 genera and between 9,000 to 10,000 species of grasses. This family comprises the most important cultivated crops like wheat, barley, rye, and rice [27]. H. chilense belongs to the Poaceae family, Triticeae tribe, genus Hordeum. Previous cytogenetic work suggested that H. chilense chromosomes are more similar to the D-than to the A-or B-genomes of wheat [28]. The phylogenetic relationships of H. chilense with respect to Triticum and Hordeum have not been studied in detail so far. For this reason, the set of EST-SSR primers that successfully amplified H. chilense and showed polymorphism was further tested for transferability to other species of the Poaceae family with the aim to investigate phylogenetic relationships.

Plant material and DNA extraction
Two accessions (H1 and H7) of H. chilense and one genotype ('Barke') of H. vulgare were used for the initial transferability analysis of 82 barley EST-derived SSR-markers. Barke was included as standard because this cultivar was used to construct the EST libraries screened to search for SSRs [29]. Total genomic DNA was isolated from young frozen leaf tissue using the CTAB method of Murray and Thompson [30] as modified by Hernandez et al. [31]. The concentration of each sample was estimated by comparing band intensity with lambda DNA digests of known concentrations under UV light, after 0.8% (w/v) agarose gel electrophoresis and ethidium bromide staining. : an initial denaturing step of 10 min at 94°C was followed by 45 cycles with denaturation at 94°C for 30 s and extension at 72°C for 30 s, respectively. The annealing temperature was decreased in 0.5°C per cycle, from 60°C in the first cycle to 55°C after the 10th cycle, and was then kept constant for the remaining 35 cycles (always 30 s). After 45 cycles, a final extension step was performed at 72°C for 5 min.

Amplification and transferability of barley EST-SSR markers
Amplification products derived from fluorescently labeled primers were resolved by capillary electrophoresis on the ABI Prism 310 Genetic Analyzer. The fragment sizes were calculated using the computer program GeneScan from the same manufacturer, by comparison with an internal size standard ( Figure 1). The presence and pattern of stuttering was locus-specific and was analyzed as an indication of correct locus transferability. For instance, Figure 1 shows the amplification of a barley trinucleotide EST-SSR, with weak -3 bp stutters. The stutters are also present in H. chilense lines H1 and H7, with a similar pattern (weak -3 bp stutters, as shown in Figure 1 pairs that showed polymorphism in H. chilense lines were tested in the other species using the same PCR conditions. To assign barley EST-SSR markers to H. chilense chromosomes, six T. aestivum/H. chilense accession H1 addition lines including a monotelodisomic 1H ch S addition, a ditelosomic addition for 2H ch alpha arm, and disomic addition lines for chromosomes 4H ch , 5H ch , 6H ch and 7H ch were used [26] (Figures 2 and 3). The addition line carrying the chromosome 3 is not available and therefore the assignment of markers to this chromosome was done using the lack of amplification in the other addition lines mentioned above.

Phylogenetic data analysis
A binary matrix was generated, where the presence or absence of each allele was coded by 1 or 0 respectively. The binary data were used to calculate the distance matrix using the Jaccard's similarity coefficient [32], because the binary information was asymmetric (the shared absence of a given allele did not contribute to genetic similarity, as no null alleles were found). The software package Phylip [33] was used to calculate the genetic relationships by neighbor-joining (NJ) analysis. The reliability and goodness of fit of the dendrogram obtained was tested by bootstrap analysis based on 1,000 permutations, followed by the program Consense module in the software package Phylip. The dendrogram (Figure 4) was constructed using the NJ method with the program TreeView [34]. The tree was rooted using B. distachyon as outgroup. For the phylogenetic analysis, hexaploid wheat allelic data were separated into its three A, B and D genome alleles, and tetraploid wheat allelic data were separated into its two A and B genome alleles (Figure 4). A second dendrogram (not shown, as the relationships are similar) was obtained only for the species belonging to the Triticeae tribe, using the UPGMA clustering method, as the divergence between these species was quite recent and therefore the molecular clock could be assumed.

Amplification and polymorphism of barley EST-SSR markers
All eighty-two sets of primer pairs from barley EST-SSR markers amplified products within the H. chilense genome. The amplification patterns of 28 EST-SSRs were not reliable though. An amplification pattern was considered as not reliable for any of the following four reasons: i) the presence of too many amplification products; ii) the presence of faint bands; iii) the molecular weight of the H. chilense amplification product was of a different size range than the H. vulgare molecular weight; iv) the H. vulgare stuttering profile was not maintained in H. chilense. Of the remaining markers, twenty-one primer pairs (26%) showed polymorphism between H1 and H7 H. chilense lines ( Figure 1) and were renamed according to their chromosomal location in H. chilense (Table 1). These twentyone primer pairs were tested in the rest of species included in this study (Table 1)

Chromosomal location of barley EST-SSR markers in H. chilense
In order to locate polymorphic barley EST-SSR markers onto the H. chilense chromosomes, a set of disomic T. aestivum-H. chilense addition lines [26] was used. All the primer pairs yielded amplification in the expected homoeologous chromosome (Figures 2 and 3). Amplification products that are present in H. chilense but absent in wheat, as well as the amplification products of different size were of particular interest, as these markers can be  Four out of 21 SSR markers amplified a SSR fragment in H. chilense, but not in wheat and 15 markers amplified PCR products of a different size range in H. chilense and wheat. Therefore, a total of 19 markers were found to be suitable for the analysis of H. chilense/wheat introgression.

Phylogenetic relationships
The potential use of the barley EST-SSR markers to infer the phylogenetic relationships among the species studied (T. aestivum, T. durum, T. urartu, T. tauschii, H. vulgare, H.  murinum, H. chilense and B. distachyon) was analyzed using the 13 EST-SSR markers that produced amplification products in all analyzed species. The NJ tree obtained (Figure 4) indicated that all the accessions were clustered according to their genome constitution. The bootstrap values are consistent and they are generally higher than 90%. The Triticum species were divided into two groups, one including the D-and B-genome of wheats, and the other including the A-genome of wheat. The Hordeum clade was separated into two clusters: one included the H. chilense accessions and the other one included H. vulgare and H. murinum species. B. distachyon was used as outgroup to root the NJ tree. The dendrogram obtained by the UPGMA clustering method (not shown) generated similar results to the NJ analysis.

Discussion
Genomic SSRs have been extensively used for mapping, genetic diversity analysis, and plant breeding, but have a lower rate of transferability across species when compared to the EST-SSR markers. Thus, the latter are a better choice for application in cross-species phylogenetic studies [35], and are also valuable tools for plant breeding, germplasm collection conservation [36], and to measure genetic diversity [37]. More than 500.000 ESTs are presently available for barley and represent an invaluable resource for the development of SSR markers [24,25]. To systematically exploit potentially useful, albeit less studied wild species, like H. chilense it would be desirable to use these EST-SSR markers as anchors to the cultivated species genome.
A set of barley genomic SSR markers [38] has been previously tested in H. chilense [11]. As expected, the level of transferability (66%) of barley EST-SSR markers found in the present work is higher than for the neutral SSRs (54%). Additionally, the level of polymorphism detected within H. chilense with the EST-SSRs (26%) is also higher than with the genomic SSRs (6%). On the other hand, the sensibly higher transfer rate to B. distachyon than to rice confirms the usefulness of Brachypo- One of the aims of this study was to infer phylogenetic relationships among the H. chilense genome, the cultivated barley genome and the wheat genomes using barley EST-SSRs (Figure 4). In the obtained dendrogram, the nodes were significantly supported by bootstrap analysis, indicating that there were subgroups that could be clearly separated.
As expected, the H. chilense genome was situated in the clade of the Hordeum species, despite the reported cytogenetic similarity with the D genome of wheat [1]. The species H. murinum (Xu-genome) was closer to H. vulgare (Igenome) than to H. chilense (H-genome). Several other studies have also grouped species of the I-and Xugenomes [41], using sequences from three nuclear regions DMC1, EF-G and ITS; and the vrs1 locus [42].
In the clade of Triticum species, the hexaploid and tetraploid wheats allelic data were separated into their genomes. A closer association was observed between the B-and D-genome and tetraploid species were closely related to hexaploid species, which is in agreement with a previous analysis based on wheat EST-SSR data [23].

Conclusion
Our study shows the utility of barley EST-SSR for the genetic analysis of H. chilense, with a remarkably high level of polymorphism within this species. It highlights a reliable and efficient way of obtaining microsatellite markers for wild relatives of a major crop. The transferred markers have shown to be useful for phylogenetic studies among the Triticeae species, and to anchor the H. chilense genome within the wheat-barley framework using Brachypodium as a root genome. The availability of additional sets of mapped EST-derived SSR markers for barley and other Triticeae genomes will assist the development of molecular maps for H. chilense and its integration into the genomic network of grass species.

Authors' contributions
AC carried out most of the molecular work and drafted the manuscript. RKV, HB, GD and PH were involved in designing and planning the work and interpreting the results. HB, RKV, GD and AG edited the manuscript critically. PH and AG conceived the study. PH coordinated the study and helped to draft the manuscript. All authors have read and approved the final manuscript.