Short Communication. Transferability of microsatellite markers located in candidate genes for wood properties between Eucalyptus species

Aim of study: To analyze the feasibility of extrapolating conclusions on wood quality genetic control between different Eucalyptus species, particularly from species with better genomic information, to those less characterized. For this purpose, the first step is to analyze the conservation and cross-transferability of microsatellites markers (SSRs) located in candidate genes. Area of study : Eucalyptus species implanted in Argentina coming from different Australian origins. Materials and methods: Twelve validated and polymorphic SSRs in candidate genes (SSR-CGs) for wood quality in E. globulus were selected for cross species amplification in six species: E. grandis, E. saligna, E. dunnii, E. viminalis, E. camaldulensis and E. tereticornis . Main results: High cross-species transferability (92% to 100%) was found for the 12 polymorphic SSRs detected in E. globulus. These markers revealed allelic diversity in nine important candidate genes: cinnamoyl CoA reductase (CCR), cellulose synthase 3 (CesA3), the transcription factor LIM1, homocysteine S-methyltransferase (HMT), shikimate kinase (SK), xyloglucan endotransglycosylase 2 (XTH2), glutathione S-transferase (GST), glutamate decarboxylase (GAD) and peroxidase (PER). Research highlights: The markers described are potentially suitable for comparative QTL mapping, molecular marker assisted breeding (MAB) and for population genetic studies across different species within the subgenus Symphyomyrtus .


Introduction
Recently, the increase of public genomic sequences and Expressed Sequence Tag (EST) data has allowed an effective approach for the identification of functional polymorphic DNA markers, such as microsatellites (SSRs). The high allelic diversity and abundance of SSRs in the forest tree species make these co-dominant molecular markers particularly useful for genetic mapping, diversity analysis and genotyping (Brondani et al., 2002;Cupertino et al., 2011). However, genomic sequence characterization including molecular marker development and application has been mainly studied in few Eucalyptus species. In accordance with the expected sequence conser-vation of transcribed regions of the genome, a significant portion of the primer pairs designed on SSR flanking regions is expected to possess a high potential for cross-transferability in related species. Furthermore, transferability of SSR markers across species of Symphyomyrtus has been proved (Glaubitz et al., 2001;Brondani et al., 2002;Faria et al., 2010Faria et al., , 2011, even in the case of the more distantly related Corymbia subgenera (Jones et al., 2001).
Many genomic studies have reported the analysis of genes expressed during wood formation and xylogenesis (reviewed by Foucart et al., 2006) and some important metabolic pathways are now well known. In Eucalyptus, several structural and regulatory candidate genes (CGs) involved in lignin biosynthesis were also identified (Paux et al., 2004). Among these genes, there are those encoding components of the common phenylpropanoid pathway, those of the monolignol specific pathway as well as lignin regulatory genes (Paux et al., 2004). However, only few studies in most widely cultivated species of Eucalyptus have analyzed CGs underlying wood quality traits using QTL approaches (Freeman et al., 2009;Thumma et al., 2010;Gion et al., 2000Gion et al., , 2011. The use of already well-established variable markers present in CGs for wood quality is an interesting approach for mapping purposes and functional diversity studies. In addition, they may be particularly useful to demonstrate if a given QTL trait such as wood quality is governed by the same genes in different species (comparative QTL or association mapping). Furthermore, cross species amplif ication of SSR-CGs loci is considered a cost-effective approach for developing microsatellite markers for new species with no or little DNA sequence information.
The present study reports the cross-transferability analysis of SSR-CG for wood quality between Eucalyptus species.

Plant material and DNA extraction
Eight trees of E. globulus were used to check reliability amplification and polymorphic status of the 21 selected primers from Acuña et al. (2012a)  Total DNA was extracted from young leaves using the CTAB method with modifications as described in Marcucci Poltri et al. (2003).

Microsatellite selection and PCR amplification
Microsatellites within CGs were selected from previous studies (Acuña et al., 2012a,b).
For validation and transferability studies, 21 SSR were identified in 19 CGs for wood quality (Acuña et al. 2012a, Online resource 1). The selection was made from the function assignment (BLASTX). To assess the putative localization of these loci, we performed a sequence BLASTN search (E-value ≤ 1e -3) against the Eucalyptus grandis genome database (version 1.1 available in Phytozome 9.1, http://www.phytozome.net/eucalyptus.php).
In addition, 8 polymorphic SSR in 7 CGs were incorporated in the transferability study (Acuña et al., 2012b).
Primers were designed to amplify these 21 regions. PCR conditions were carried out in a final volume of 12 µl with 20 ng of genomic DNA, 0.25 µM of each primer (Alpha DNA, Canada), 2mM MgCl 2 , 0.2 mM of each dNTP, 1X reaction buffer and 1U Platinum Taq polymerase (Invitrogen). A f inal concentration of 3 mM MgCl 2 was used for the transferability analyses.
Amplifications were performed following a denaturation step of 5 min at 94°C, 35 cycles of 1 min at 94°C, 1 min. at annealing temperature and 1 min at 72°C. The final extension step was of 10 min at 72°C. Details of primers sequences, SSR location, amplification conditions, and product sizes are described in Table 1. Fluorescent dye-labeled forward primers (Alpha DNA, Canada) were used and fragments were separated on an ABI3100 Genetic Analyzer (Applied Biosystems, USA). Allele assignments were made by size comparison with the standard allelic ladders, using the GenMapper ID software provided by Applied Biosystems (USA).

Genetic Analyses
The number and frequency of alleles, as well as the observed heterozygosity (H o ), were determined using the GenAlEx 6.4 program (Peakall & Smouse, 2006).
The polymorphism information content (PIC) of each marker was calculated according to Botstein et al. (1980). Shared Allele Distance (DAS), defined as one minus half the average number of shared alleles per locus, was implemented between individuals using the software POPULATIONS 1.2.28 (Langella, 2002).
Cluster analysis was implemented based on distance matrices using the unweighted pair group method arithmetic average (UPGMA), and the corresponding phenogram was constructed.

Results and discussion
From the 21 SSR-CGs involved in phenylpropanoid biosynthesis, cellulose biosynthetic process and me-thionine metabolism that has not been previously validated, nine markers (43%) generated reproducible and reliable amplicon patterns in the E. globulus sample tested. From these nine markers, five markers were monomorphic and four polymorphic in this small sample. Furthermore, these loci were physically aligned to a unique position on the chromosome scaffolds 3, 7, 8 and 11 (Table 1). The polymorphism rate was similar to those described by Torales et al. (2012), in Nothofagus (19%) but lower than those described by Faria et al. (2011), (39%) and Acuña et al. (2012a), (46%) in Eucalyptus.
In this work, we also assessed the cross transferability of 12 SSR-CG by using new and previously validated primers: four polymorphic primers validated here and other eight primers previously validated in Acuña et al., (2012b). A high cross transferability (92% to 100%) was found in E. grandis, E. saligna, E. dunnii, E. viminalis, E. tereticornis and E. camaldulensis. As the primers were designed from genic regions, it is reasonable that they have a high potential for cross-transferability in related species. In fact, in a recent study, similar cross-species transferability of SSR markers was already reported for species of the subgenus Symphyomyrtus (Zhou et al., 2013).
Nine out of 12 loci showed polymorphic patterns within all species analyzed, while XTH2, HMTe and HMTi loci were monomorphic in some species (Table 2). However, their classification as monomorphic loci is still preliminary because of the relatively limited number of species and trees that were tested per species. Furthermore, average values of number of alleles 508 C. V. Acuña et al. / Forest Systems (2014) 23 (3): 506-512 However, the values found in the present study are lower than those described by Faria et al., (2010), also in E. grandis. These lower results could be explained because of the difference in selection criteria. While Farias et al., (2010) based their selection on polymorphism level, we selected the loci according to their putative function in wood quality development.
In general, in the current study the average number of alleles (N a ) as well as the average observed heterozygosities (H o ) and PICs were higher in SSRs located within predicted promoter or intron regions (6.8, 0.68 and 0.75, respectively) than in exons (4.3, 0.48 and 0.51, respectively). Analyzing across species, we found the largest number of alleles for E. tereticornis, E. viminalis and E. camaldulensis. The SSRs charac-terized in the current study may be useful for population genetic studies in these species. This is particularly useful, since a few SSR markers have been reported until now for these species (Yasodha et al., 2008;Acuña et al., 2012a).
The 12 SSR-CG markers distinctively fingerprinted the 52 genotypes analyzed, allowing the construction of individual molecular identification patterns. This was corroborated with a genetic distance analysis between all 52 individuals that showed a mean of 0.78 and only 4 pairs of trees with distances lower than 0.25. These results reflect the high resolution capacity of microsatellites.
Estimates of average shared allele distance (DAS) among individuals within species ranged from 0.48 to 510 C. V. Acuña et al. / Forest Systems (2014) Faria et al. (2011). As expected, these values increased when considering DAS between individuals of different species (82% of pairwise data were higher than 0.7). Finally, the cluster analysis showed that this set of markers allowed a clear-cut discrimination among all individuals, reflecting a high resolution capacity of microsatellites (Fig. 1). For E. globulus, E. viminalis and E. grandis, the individuals of the same species grouped together. Other clusters observed were congruent with taxonomic relationships described by Steane et al. (2011): individuals of E. tereticornis and E. camaldulensis (section Exsertaria) grouped together with E. saligna and E. grandis (section Latoangulatae).

Conclusion
High cross species transferability (92% to 100%) of this polymorphic SSR-CGs was demonstrated. Potentially, these markers may contribute to the verification of synteny and collinearity between different Eucalyptus maps. Furthermore, they would also allow the validation of gene and QTL positions in multiple pedigrees in the botanical sections to which most of the commercially planted eucalypt species belong: Maidenaria, Exsertaria, and Latoangulatae.