Cross transferability of SSRs to five species of Araucariaceae : a useful tool for population genetic studies in Araucaria araucana

Cross-species amplif ication of microsatellites is a common procedure to obtain suitable markers to be used in population genetic studies. Primers designed for one (source) species are used to amplify homologous loci in related (target) species. It is expected that phylogenetically close species will share a higher proportion of markers, and genetic distance could be a useful parameter to predict successful transferability between different taxonomic groups. We analyzed twenty-two primer pairs developed for Araucaria angustifolia (Bertol.) Kuntze in five target species of the Araucariaceae family. The results were summarized in vectors of presence and absence of bands and compared through the Jaccard similarity index. Using the sequences of eight published genes, genetic distances between pairs of species were estimated and related to transferability rate using Pearson correlations. Successful transfer rate ranged from 31.8 to 77.3%, being these among the highest reported for plants. The highest transfer rate was observed between the South American species. The transferability was confirmed sequencing seven fragments amplified in A. araucana (Molina) K. Koch, and using the best five to estimate genetic diversity parameters in a natural population of this Andean coniferous. Key word: cross species amplification; pairwise genetic distance; conservation of microsatellite loci; Pehuén.


Introduction
Microsatellites or Simple Sequence Repeats (SSRs) are among the most useful DNA markers for studying population genetic structure and dynamics (Zhang and Hewitt, 2003).Although their development has become more accessible in recent years, the cost and effort to obtain SSRs is still significant.Cross-species amplification of SSRs is therefore a common practice (e.g.Kayser et al., 1996;Kijas et al., 1995;Kupper et al., 2008;Lin et al., 2008).It is assumed that the transferability success depends on the extent of sequence conservation in the primer sites flanking the microsatellite loci and the stability of these sequences during evolution (Zhang and Hewitt, 2003).Therefore, it is expected that the chance of successful cross-species amplification is inversely related to the genetic distance between the species (Zucchi et al., 2002).
Araucariaceae family, whose origin was estimated to be 308 ± 53 Ma (Late Carboniferous) (Liu et al., 2009), includes three genera: Araucaria, Agathis and Wollemia, today restricted to the Southern hemisphere (Hill and Scriven, 1995).Most species occur in the ecozones of Indomalaya and Australasia.Two species are distributed in South America: Araucaria angustifolia and Araucaria araucana (both belonging to the section Araucaria). A. araucana (Pehuén) is endemic to the South American temperate forests.Its distribution range along the Andes spans between 37°27' S and 40°03' S latitude and it is also found in two isolated populations in the Chilean coastal mountains.It is mostly dioecious; pollen is dispersed by wind and seeds by gravity and small rodents.A. araucana is classified under IUCN (International Union for Conservation of Nature and Natural Resources) guidelines as vulnerable (Farjon and Page, 1999) and currently it is officially protected in CITES (Convention on International Trade in Endangered Species of Wild Fauna and Flora).In this species, several studies on genetic diversity were performed, using isozymes (Gallo et al., 2004;Ruiz et al., 2007), RAPD (Bekessy et al., 2002) and chloroplast DNA (Marchelli et al., 2010) markers.However, for gene flow and fine-scale genetic structure it is necessary to have highly polymorphic co-dominant genetic markers like SSRs.
Nuclear microsatellite loci were developed for different species within Araucariaceae and their transferability to other species within the family was evaluated (Robertson et al., 2004;Scott et al., 2003;Salgueiro et al., 2005;Schmidt et al., 2007).Salgueiro et al. (2005) evaluated transferability of all the microsatellite primers developed from Australasian species of the family available until that moment, to both South American species, finding only two successful transferences.Accordingly, they developed six new SSR loci that were reported as useful to A. araucana.However, and in agreement with the observations of Scott (2004) in other species of the family, we observed a significant failure in some of these primers when they were screened widely in the Andean species.Therefore, a proper set of highly polymorphic microsatellite markers for gene flow studies in A. araucana does not exist.For this reason, an effort to obtain these markers through cross-species amplification of microsatellites developed in A. angustifolia (Schmidt et al., 2007) was made.
Additionally, this set of primers was tested in other phylogenetically more distant species within the Araucariaceae family, also to evaluate the relationship between cross-species amplification using genetic distance.For this purpose, four species of the genus Araucaria (Araucaria angustifolia, A. bidwillli (Molina) K. Koch, A. cunninghamii Aiton ex D. Don, A. heterophylla (Salisb.Franco) and one of the genus Agathis (Agathis alba Rumph.ex Jeffrey) were considered.According to the phylogenetic relationships, the expectation is that most of the microsatellite loci developed in Araucaria angustifolia would be successfully transferred to Araucaria bidwilli and Araucaria araucana, with a similar degree of polymorphism in the later.For the other phylogenetically more distant species, a lower transferability rate is expected.To test this hypothesis, the amplification success of each microsatellite was evaluated in relation to the genetic distance between the target and the source species.Genetic distance between species was estimated using two mitochondrial (coxI and atpI), four chloroplast (rbcL, matK, rps4 and cp16S) and two ribosomal genes (18S and 26S).We demonstrate that there is not a significant correlation between cross-amplification success and genetic distance among these members of the Araucariaceae family.

Sample collection and DNA extraction
Six Araucariaceae species were studied: Agathis alba, Araucaria angustifolia, A. araucana, A. bidwillli, A. cunninghamii and A. heterophylla.The specimens were collected from different gardens in Argentina, and from a natural population of Araucaria araucana (Tromen 39°37' S, 71°20' W, 984 m.a.s.l.).Leaves from three individuals per species were sampled to test transferability of primers, and 79 individuals from one population of A. araucana were collected in order to evaluate polymorphism in the species.All the samples were store at -80°C until DNA extraction.
Total genomic DNA was extracted following the protocol by Stefenon et al. (2004), after grinding the leaves to a f ine powder with a mixer mill (Retsch, Germany).DNA concentration was estimated either on 0.8% agarose gels or using a photometer.

Microsatellites analysis
Twenty-two primer pairs isolated from A. angustifolia (Schmidt et al., 2007;Salgueiro et al., 2005) were tested in the five target species.Three individuals per species were analyzed, except for A. araucaria where the number of individuals varied between 3 and 10, depending on the primer.Polymerase chain reaction (PCR) was carried out in a final volume of 13 µl, with 1X GoTaq reaction Buffer (Promega), 1,53 mM of polyvinylpyrrolidone (PVP, Sigma), 300 µM of each dNTPs (Invitrogen), 0,85 U of Taq Polymerase (GoTaq, Promega), 1 mM of MgCl 2 , 0.5 µM of each primer and 5 ng of DNA.PCR thermal profile for all species was as follows: denaturation at 94°C for 4 min followed by 35 cycles at 94°C for 30s, 52-60°C for 30s, 72°C for 30s and final extension at 72°C for 10 minutes.PCRs were performed using a My Cycler thermal cycler (BIORAD).Lack of amplification under these experimental conditions was recorded as transferability failure to the target species, and no additional PCR optimization experiments were performed.
On the other hand, optimization experiments of the 22 primer pairs were carried out in A. araucana, testing factorial combinations of three different annealing temperatures, T a (T a used in source species, T a -2°C, T a + 2°C), and different MgCl 2 (1.5, 2.5 and 3.5 mM) and DNA concentrations (5, 10, 15, 20, 25 and 30 ng).The optimal combination for each primer is reported in Table 1.
PCR products were ran on 2% agarose gels with 0.5 X TBE buffer at 60 V for 10 min and at 120 V for 90 min, and visualized under blue light after staining with Syber Safe (Invitrogen).Although the amplification was ranked according to the quality of the fragments and the presence of non-specific amplification products, transference was considered successful when amplicons of the expected size range were visualized on agarose gels.
The amplified fragments obtained with the working primers were then ran on a 6% standard denaturing polyacrylamide gel.PCR products were mixed with 95% formamide, 0.05% bromophenol blue and 0.05% xylene cyanol and denatured at 94°C for 5 min.Gels were ran at 80 Watt for 3 h and silver stained following the protocol by Bassam et al. (1991).Slippage patterns in the polyacrylamide gel are indicative of fragments of small repetitive units and therefore hint to microsatellite regions (Dowling et al., 1996).
In order to confirm the presence of microsatellites, PCR products obtained with seven successful pairs of primers in A. araucana were sequenced (three individuals per primer).The sequence was obtained by direct sequencing (forward and reverse) using the BigDye chemistry (Applied Biosystems, Foster City, CA) and analyzed on an ABI3130XL automatic sequencer (Genomic Unit, CNIA INTA Castelar, facilities).
In A. araucana, the screening for polymorphism was done only for those primers showing a high amplification quality or displaying a high number of alleles in the source species.Between 52 and 75 individuals were amplif ied with f ive fluorescently endlabeled primers and PCR was carried out under the conditions described above.The samples were run on a MegaBace 1000 (GE Healthcare) automatic sequencer and electropherograms analyzed using the MegaBACE Fragment Profiler version 1.2 software (GE Healthcare).

Data analysis
The results for the 22 loci were summarized for each species in a vector of 0 (no amplif ication) and 1 (successful amplification).Vectors were compared through the Jaccard similarity coefficient using the software NTSYS (Rohlf, 2001).This coefficient assigns more weight to the double presence of a band than to its double absence.
Uncorrected genetic distances (p-distance, Kumar et al., 1993) between target and source species were estimated at 18S, 26S, rbcL, matK, rps4, cp16S, coxI and atpI genes.Sequences were retrieved from GenBank, edited and aligned with MEGA (Tamura et al., 2007).Additional information on microsatellite cross-species amplification was gathered from studies in other Araucariaceae species.The data were included in the comparison when (1) the source species of the microsatellite loci and the amplification success were clearly specified in the target species, and (2) 18S, 26S, rbcL, matK, rps4, cp16S, coxI and atpI sequences were available for both source and target species.Overall, information regarding cross-species microsatellite transferability success for loci isolated from A. angustifolia, A. bidwilli and A. cunninghamii, available before the publication by Schmidt et al. (2007), was compiled and presented in Table 2 (upper part), together with our results for the 22 primers developed by Schmidt et al. (2007).The percentage of loci successfully transferred was calculated for each pair source-target species (CSA% Cross Species Amplif ication; Table 2).We classif ied the transference according to the quality of the amplified product as 1: excellent amplification product, 2: no specific amplification, 3: multiple bands, smear or faint amplification product, 4: no amplification.
Genetic distances (p) and amplification rates were compared by Pearson's correlation (r) using R 2.10.0 (R-Development-core-team.2008).Matrices of genetic distances (p) and Jaccard distances (dissimilarity, Jd) were compared by Mantel test using the program NTSYS (Rohlf, 2001).To confirm the presence of the microsatellite stretches in A. araucana, the sequences were aligned and compared against the sequences reported for the source species (obtained from NBCI Genbank), using the program Bioedit (Hall, 1999).
In A. araucana, number of alleles per locus (N A ), effective number of alleles (N e ), exclusion probability (P E ) [following equation 2a in Jamieson and Taylor (1997)], expected (H e ) and observed (H o ) heterozygosities (Nei, 1971)  CSA %: microsatellite Cross-Species Amplification success between source species and target species (according to the quality as defined in Table 1).Origins of data used to estimate the percentages are specified in the column references.NM: number of microsatellite loci analyzed in each referenced work.N: number of analyzed samples.
expectations were estimated using GenAlEx 6.3 (Peakall and Smouse 2006).The same genetic parameters were calculated for three loci in A. angustifolia using the allele frequencies reported by Patreze and Tsai (2010).
For the other loci, allele frequencies were not available.The presence of null alleles at each locus was verified with MICRO-CHECKER version 2.2.0.3 (Van Oosterhout et al., 2004).

Results
Transferability was successful for 7, 7, 10, 10 and 17 SSRs in A. cunninghamii, Agathis alba, A. bidwilli,A. heterophylla and A. araucana,respectively (Table 2,upper part).Cases classified in category 3 (faint amplification products) were considered as a successful transference.On the other hand, the screening of polymorphisms was performed using only the primers classified in category 1. Jaccard similarity coefficients between target species and A. angustifolia (source species) were 0.368, 0.421, 0.450, 0.450 and 0.800 for A. cunninghamii, Agathis alba, A. bidwilli, A. heterophylla and A. araucana, respectively.The relationships between p genetic distances for the eight genes and amplification rates were non significant (Table 2, lower part).No significant correlations between p genetic distance and Jaccard distances (Jd) for any of the eight genes studied were observed.The highest correlation was obtained for the gene 18S (r = 0.393, p = 0.074).
In A. araucana, loci Aang 01, Aang 03, Aang 18, Ag 23 and Ag 56 showed high quality of amplification; the locus Aang 15 displayed a high number of alleles in the source species and in the Andean species and therefore an effort was made to improve the amplification quality (Table 1).These six loci, together with locus CRCAc 1 (reported as transferable by Salgueiro et al., 2005) were sequenced.All the sequenced fragments contained the same repetitive motif that was reported originally in the source species (See Annex 1 for details).Due to the difficulty in obtaining adequate PCR products for primers of quality 3 (smear, faint or multiple bands) we limited the screening for polymorphism to those primers of high quality, except when the numbers of alleles obtained were high.In A. araucana, the f ive successfully tested microsatellites (Ag23, Ag56, Aang15, Aang18 and CRCAc1) segregated in a Mendelian way, as evidenced by comparing embryos with the parental genotype assessed from megagametophytes (data not shown).The number of alleles ranged between 3 and 22 and the effective number of alleles between 1.45 and 8.73.Significant deviations from Hardy-Weinberg expectations were detected at three loci (Ag 23, Aang 15 and CRCAc1).At locus Aang 15, significant evidence for the presence of null alleles (Fisher combined probability test p < 0.001) was observed.Observed and expected heterozygosity ranged from 0.307 to 0.685 and from 0.282 to 0.890, respectively (Table 3).Exclusion probability (Jamieson & Taylor 1997) was 0.816.

Discussion
For a successful cross-species amplification of microsatellite markers, the repeat sequence and the flanking regions containing the selected primer sites must be conserved across taxa.Therefore a higher ge-  et al., 2005et al., (from Scott et al., 2003) ) nomic homology is likely to translate into greater conservation of SSR flanking regions and, as a result, in higher transferability of primer pairs (Rossetto, 2001).Studies within the Araucariaceae family showed a very low genetic variability among and within species (Peakall et al., 2003;Scott et al., 2005).Moreover, after sequencing, similar structure of the microsatellite stretches and a high sequence conservation of the flanking regions was observed among several species of Araucariaceae (Scott et al., 2003).These results suggest the possibility of a successful transferability of SSRs within the family.
In our study, successful cross-species transferability varied between 31.8 to 77.3% (CSA % in Table 2), similar to what is reported in others Araucariaceae (45.5-100%,Scott, 2004).Although the amplification rate decreased with phylogenetic distance, no significant correlations were detected.The highest percentage of transferability occurred among the American species of the genus Araucaria (A. angustifolia and A. araucana), in spite of the closer phylogenetic relationships between A. angustifolia and A. bidwilli (Liu et al., 2009).In general, the primer transferability from the American to Australian and New Zealand species of the family (or vice versa) was among the lowest.The highest transferability rates between species were observed within each tribe of the Araucaria genus: 77.3-100% and 81.3% within tribe Columbea (Araucaria) and Eutacta, respectively (Salgueiro et al., 2005 andScott et al, 2003).On the other hand, the genus Agathis (which originated during the late Jurassic/early Cretaceous) showed a higher degree of transferability (36.4% in our study, 45.5% in Scott 2004) compared to other families with lower divergence time: 22-35.3% in Quercus (Isagi and Suhandono, 1997;Steinkellner et al., 1997), 10.2% in Picea (Ven and McNicol, 1996), 20% in Pinus (Fisher et al., 1998), 38.3% in Melaleuca (Rossetto et al., 2000) and 20% in Eucalyptus (Brondani et al., 1998).
The presence of the repetitive motifs was verified by sequencing, concluding than these microsatellite loci are orthologous in A. araucana (see Annex 1).The homology of sequences between the analyzed species was variable, although the same repetitive motif was identified in each case.Locus Aang 15 presented a high number of alleles in A. araucana, although the effective number of alleles was similar to that in the source species A. angustifolia (8.73 and 8.83 respectively, Table 3).Salgueiro et al. (2005) reported a complete transferability of six markers developed for A. angustifolia testing six individuals of A. araucana.However, in our study only two of them could be amplified, with a high rate of amplification failure for the rest of the primers when screened on a large sample size.Similar results were obtained by Scott (2004) in A. cunninghamii where some loci required significant optimization of the PCR conditions.The amplification failure could be due to point mutations (SNPs) in the annealing site, which could therefore reduce the transferability rate.On the other hand, technical limitations reducing SSR transferability rate in A. araucana cannot be completely ruled out.In fact, although the south American Araucarias share a common ancestor (estimated divergence times by fossil records is 98.9-142 MYA), the process of speciation restricted A. araucana to the cold Andes region and A. angustifolia to the moist regions of the southern Brazilian highlands (Stefenon, 2007).These differences in the ecological niches are reflected in their types and contents of foliar terpenes (Schmeda-Hirschmann et al., 2005).Csaikl et al. (1998) reported that terpenes, polyphenols and polysaccharides are abundant in the foliage of perennials, and are coextracted with DNA.The higher amplification failure and the lower number of loci with high amplification quality in A. araucana could be in part due to the presence of secondary metabolites that inhibit amplification, and the same could occur within other species of this family.
Sequencing these fragments allowed us to confirm unequivocally the presence of the microsatellite repeats in A. araucana and to verify their structure.In those species of the family Araucariaceae in which transferability was verified, it would be important to determine if polymorphism exists and if effectively the repeats are conserved, before using them as molecular markers for population genetics studies.In the case of A. araucana, we obtained a set of five microsatellite primers showing a high exclusion probability suitable for population genetic analysis that are currently being applied in pollen and seed flow studies.
primers for this study, to A. Aparicio for helping with the statistics, V. Arana and V. El Mujtar for their assistance with sequences analysis.We are indebted to Maria Marta Azpilicueta for discussion on data analysis, laboratory assistance and for making constructive comments on the manuscript.This study was partially financed by INTA (National Institute for Agricultural Technology) through the project AEGR140.The study was also partially supported by CONICET (PIP 5451 ) denotes PCR product size in target species.Q: quality of the amplified product (1: excellent amplification product, 2: no specific amplification, 3: multiple bands and smear or faint amplification product, 4: no amplification).N: number of individuals tested.T a : annealing temperature.Amount of DNA and concentration of MgCl TGGAGT T T AG G TGGATAAAC AT GTAT TGGA TAAGCACAT T GT CCCAT A. araucana TGGAGT T T AG G TGGATAAAC AT GTAT TGGA TAAGCACAT T AT GAT T AT AC T C TTAACCCT AT T C TA TATC AAAATT TG TA T TC T C TCT C T A. araucaria AT GAT T AT AC T C TTAACCCT AT GCCA TAGC AAA T TT TG TA T T C T C TCT C T . . . .| . . . .| . . . .| . . . .| . . . .| . . . .| . . . .| . . . .| . . . .| . . . .| 110 120 130 140 150 A. angustifolia C T C TC T CT C T C T C-----AT CAAAT T ACCT T C T TGTATCTG AT TGGCCAAC A. araucaria C T C TC T CT C T C T CT CT CATC AAAT TACCTT C T T GTA TC TG AT TGGCCAAC

Table 1 .
Cross transferability of SSRs in Araucaria species305 Details of cross-species amplification of 22 nuclear microsatellites developed for Araucaria angustifolia (source species) and tested in five Araucariaceae species (target species)

Table 2 .
and deviations from Hardy Weinberg Cross transferability of SSRs in Araucaria species 307 Upper part.
p: genetic distance for 18S, 26S, rbcL, matK, rps4, cp16S, coxI and atpI gene sequences respectively.Lower part.Pearson correlation's coefficients and respective p values between genetic distance at each gene and CSA %

Table 3 .
Loci transferred from Araucaria angustifolia to Araucaria araucana, and estimated genetic parameters ). A. C. Moreno was supported by a fellowship for graduate students of CONICET.Comparison of sequences between the source species (A.angustifolia and A. cunnighamii) and A. araucana (target species).Repetitive motifs are underlined.Locus name is mentioned before each alignment of sequences Comparison of sequences between the source species (A.angustifolia and A. cunnighamii) and A. araucana (target species).Repetitive motifs are underlined.Locus name is mentioned before each alignment of sequences