INTRASPECIFIC VARIATIONS STUDIED BY ISSR AND IRAP MARKERS IN MASTIC TREE (Pistacia lentiscus L.) FROM TURKEY

In this study, intra-specific variations in naturally growing and cultivated mastic tree (Pistacia lentiscus L.) samples obtained from western parts of Turkey were examined using ISSR and IRAP marker techniques. Samples from Crete and Chios were also included in the study. Morphological measurements of some leaf characteristics of the samples were performed and the measured data was evaluated statistically with a Pearson Correlation analysis to reveal the correlations between character pairs. ISSR primers produced 81 bands between 161-1884bp with 96.3% polymorphism and IRAP primers produced 72 bands between 124-2027bp with 91.67% polymorphism. Polymorphism information content (PIC) values were 0.458 and 0.418 for ISSR and IRAP, respectively. Genetic similarity matrix was examined with Jaccard’s coefficient. Maximum similarity was found between the Cretan samples (LG2 and LG3) with the ISSR analysis (0.933) and between L25A (C1, Bodrum) and L29A (C1, Milas) with the IRAP analysis (0.593). Unweighted pair group method with arithmetic mean (UPGMA) dendrogram was divided into 12 and 4 groups by ISSR and IRAP methods, respectively. Specimens were segregated on 3 main different clusters by the Principal Component Analysis (PCA) based on the combined marker systems. The results showed that P. lentiscus has very high ratios of intraspecific variation. The present work is an original study in terms of large sampling including wild genotypes, cultivated specimen, Chios and Cretan varieties, use of ISSR and IRAP combination, determination of relations between culture and wild genotypes and the use of Bagy-1 retrotransposons in intraspecific polymorphism. This study may be considered as a reference study for studies on gene pools of P. lentiscus and phylogenetic relationships within the species and may contribute to species concept and agricultural breeding programs.


Introduction
The genus Pistacia L. within the family Anacardiaceae is represented with 11 species worldwide and with six naturally growing species in Turkey (Stevens 2008, Kafkas & Perl-Treves 2001, Kokwaro & Gillet 1980, Whitehouse 1957, Yaltırık 1967, Zohary 1952).Members of the genus grow naturally in various areas of the northern hemisphere including the Middle East, Canary Islands and the Mediterranean region (Ak & Parlakçı 2009).Pistacia genus is believed to have originated in Central Asia 80 million years ago (AL-Saghir 2010, Parfitt & Badenes 1997).Mastic tree (lentisk) (Pistacia lentiscus L.) is an economically P. Turhan-Serttaş & T. Özcan important species of the genus in terms of mastic resin which has a wide spectrum of biomedical usage and cultivated for its aromatic resin.As an evergreen dioecious maquis element with a height ranging from 1 to 5m and a strong smell of resin, P. lentiscus bears fleshy drupe with an initial red colour which becomes black after ripening.The leaves are compound paripinnate, alternate, and leathery.The high ecological tolerance of the species allows it to resists heavy frosts and drought (Correia & Catarino 1994) and to grow well in all types of soils ranging from limestone areas to saline environments around sea (Zohary 1952).Its resin, known as mastic, is harvested as a spice from cultivated mastic trees and used in food industry, cosmetics and medicine with therapeutic effects such as gastrointestinal ailments, and antibacterial and antifungal properties.Mastic resin is reported to be effective in inhibition of cell lines of some cancer types including colon (Balan et al. 2007), prostate (He et al. 2007), and erytrolosemi (Loutrari et al. 2006).Analysing gene pools and conservation of genetic resources of P. lentiscus are important strategies for selection of high yield genotypes adapted to various environmental conditions and for increasing product potential in breeding programmes.Pistacia has formerly been divided into four groups as Lenticella, Eu-Lentiscus, Butmela and Eu-terebinthus including 11 species in total based on some morphological observations of Zohary (1952) and Whitehouse (1957).In the first molecular study based on chloroplast DNA, two groups, Terebinthus and Lentiscus, represented by deciduous and evergreen plants were suggested to be placed in Pistacia (Parfitt & Badenes 1997).The group Terebinthus was also supported by some molecular studies carried on Mediterranean Pistacia species (Golan-Goldrish et al. 2004, Kafkas and Perl-Treves 2001, Kafkas & Perl-Treves 2002, Kafkas 2006).Some morphological characters such as features of rachis and compound leaves, shape, colors and venation of the leaflets, shape of fruits are used in taxonomical delimitations of Pistacia species in Turkey (Yaltırık 1967).Some studies were performed on Turkish Pistacia species based on random amplified polymorphic DNA (RAPD) (Kafkas & Perl-Treves 2002) and amplified fragment length polymorphism (AFLP) markers (Kafkas 2006) explaining some taxonomical relations of the species and suggesting some nomenclatural combinations.In a recent work carried on P. Lenticus in Turkey, Inter Simple Sequence Repeat (ISSR), Simple Sequence Repeat (SSR), RAPD and Internal Transcribed Spacer (ITS) markers were used to evaluate patterns of genetic variation and phylogenetic relationships in 24 wild-type mastic trees (Abuduli et al. 2016).The ISSR results indicated that male and female genotypes were distinctly separated from each other and that ISSR markers were useful for analysis of intraspecific variations in mastic trees.
ISSRs are defined as inter-gene regions of microsatellites and ISSR markers are used as molecular markers in polymorphism studies.Microsatellites occuring at thousands of locations within a genome are widely used for DNA profiling, genetic diversity, genetic linkage analysis and marker assisted selection (MAS) to locate a gene or a mutation responsible for a given trait.They are also used, owing to their high polymorphism rates, in phylogeographic studies (Nagy et al. 2003) and to measure the levels of relatedness between infraspecific taxa, groups and individuals (Fritz et al. 2005).As an alternative valuable retrotransposon-based marker, interretrotransposon amplified polymorphism (IRAP) markers are also used to determine genotypes, measure diversity, establish taxonomical relations or reconstruct phylogeny (Kalendar et al. 1999).Retrotransposons which are mobile genetic elements can induce mutations by inserting near or within genes.The copy number of retrotransposons exhibit diagnostic patterns among closely related plant taxa (Tenaillon et al. 2011) and are useful for determining genetic diversity within a species (Kolano et al. 2013) in addition to genomic changes associated with retrotransposon activity in abiotic stress conditions (Fan et al. 2014).Retrotransposons are divided into two groups according to the presence or absence of long terminal repeats (LTR).The LTR retrotransposons were well analyzed at different taxonomical categories of plants (Park et al. 2007, Ma et al. 2008).
In this study, intra-specific variations in 35 samples of naturally growing and culture varieties of P. lentiscus obtained from various localities in Turkey, Chios and Crete were examined using ISSR and IRAP marker techniques in order to provide data for elucidating the gene pool of the species in the sampling region, contribute its revision, taxonomical delimitation and its phylogeographical relations.For phylogenetic analysis, dendrograms were generated based on ISSR and IRAP data and the marker results were evaluated by PCA analysis.Morphological measurements of the collected specimens were also carried out to determine similarity coefficients of phenotypic characters at population level.

Plant material
Pistacia lentiscus specimens were collected from 31 native populations distributed in İstanbul (A2(E)), İzmir (B1), Muğla and Aydın (C1 and C2) according to the grid system of Turkey.Four specimens collected from Chios (1) and Cretan (3) were also included in the study (Table 1).The collected specimens were prepared as herbarium materials and are kept in ISTF (Herbarium of İstanbul University Faculty of Sciences).Leaf samples of the collected specimens were transported to the laboratory in polypropylene bags and kept at -80ºC conditions until the analysis.

Morphological analysis
The lengths of leaflet tips, widths and lengths of rachis, leaf and leaflets were measured and leaflet numbers and the shapes of leaflets were determined using binocular stereomicroscope.Each measurement was performed as five replicates for each character.A variance analysis (ANOVA) was performed with the obtained data to determine phenotypically differences.The Pearson correlation test was performed to determine the relationships between measured characters.

DNA isolations
Genomic DNA isolations of P. lentiscus specimens were performed by the CTAB method of Lodhi et al. (1994) with minor modifications.The concentration and purity of the genomic DNA samples were evaluated in the Nanodrop 2000c instrument.Qualities of the genomic DNAs were analyzed by running on 0.8% agarose gel using 1xTBE buffer and ethidium bromide (EtBr).Scanning of the gels was carried out under UV (GELIANCE 200 Imaging System with GeneSnap software).

PCR Amplification of ISSR and IRAP
Polymerase chain reaction (PCR) amplifications of the DNA samples were carried out in Thermo Scientific Arktik Thermal Cycler using 12 ISSR and 14 IRAP primers (Table 2).The primers showing clear bands on gel images were selected.The ISSR amplifications were analyzed in volumes of 20μl including 1xPCR buffer, 2.5mM MgCl2, 0.4mM dNTP mix (dATTP, dGTP, dCTP and dTTP), 1U Taq DNA polymerase, 0.4mM for each ISSR primer, 50ng DNA sample and DNase/RNase free water.The PCR reactions for ISSR markers were programmed with the following conditions: after initial denaturation of DNA at 95°C for 5min, 35 cycles of denaturation at 95°C for 30sec, annealing (Ta°C is different for each primer) (Table 2) for 1min, 72°C for 1min and final extension at 72°C for 5min.The IRAP

Evaluation of the datasets
The "Thermo Scientific myImage Analysis v2.0" software was used to determine the band molecular sizes using the gel images.The band fragments were visually scored for presence (1) or absence (0) on their gel patterns (additional data are given at the journal's web page as Supplemantary Material Table 6).The polymorphism information content (PIC) value was calculated according to the formula; where Pi represents the frequency of polymorphic bands present for a primer (Bhat 2002).The similarity coefficient matrix of the analyses was made based on the "Jaccard" similarity formula (Jaccard 1908) and was generated with the PASW18 software.The results were evaluated by constructing dendrograms for ISSR and IRAP according to the Jaccard's similarity coefficients.In order to determine phylogenetic relationships, unweighted pair group method with arithmetic mean (UPGMA) algorithm was produced using the XLSTAT package program.The distribution of the genotypes in the P. lentiscus gene pool was assessed by the principle component analysis (PCA) based on ISSR and IRAP data.The PCA analyses were also performed using the XLSTAT package program.

Morphological analysis
The results of the morphological measurements (see Table 3) showed that the minimum and maximum values were 0.31mm and 1.42mm for the tip length of leaflets.The minimum rachis length was measured as 1.93cm for L28A and the maximum as 5.71cm for LZB and the rachis width ranged from 1.10mm (L16B) to 2.69mm (L11A).The lowest leaf size was measured as 3.66cm for L28A and the highest value as 7.68 cm for LZB.The average value of the leaflet lenght in the population was calculated as 2.73cm (Table 3).The results of the Pearson correlation analysis were given in Table 4.A high correlation level (p<0.01) was determined between the pairs of the morphological characters, and the highest correlation (r=0.901) was measured between leaf width and leaflet size.

ISSR and IRAP polymorphism analysis
Although a total of 12 different primers were used for ISSR analysis, 9 of them were included in the band analysis.A total of 81 bands and 1444 amplicons were obtained between 161 bp and 1884 bp per ISSR primer and 3 of them were monomorphic band (Fig. 1).The lowest and the highest PIC values were 0.368 and 0.495, respectively.The mean PIC value was 0.458 and the polymorphism rate was found to be 96.3%(Table 5).
14 different primers were tested in the IRAP-PCR.Bagy-1, BARE-1 and Sukkula LTR-retrotransposon primers were determined after the PCRs.Bagy-1 retrotransposon primer (IRAP9), LTR4 primer (IRAP12) and LTR2 primer (IRAP14) were selected based on the clearly produced band profiles in agarose gel electrophoresis, and the other primers were not included in the assays since they were determined to be monomorphic.The band results of the IRAP-PCR revealed a total of 737 amplicons from 72 band fragments between 124 bp and 2027 bp (Fig. 2).The mean PIC value for IRAP marker assay was found to be 0.418 (Table 5).

Genetic similarity analysis
The percentages of similarity obtained as a result of whole ISSR and IRAP analyses were calculated with Jaccard's coefficient.According to ISSR datasets (additional data are given at the journal's web page as Supplemantary Material Table 7), similarity ratios varied between 0.080 and 0.933.The lowest similarity was found to be 8% between LBB and L33A, and the highest (93%) was between LG2 and LG3.The lowest polymorphism was found in LBB genotype with an average value of 0.127 and the highest polymorphism was in L28B genotype with an average value of 0.628.The average values of the highest and lowest polymorphism values were calculated as 0.712 and 0.125, respectively.According to IRAP datasets (additional data are given at the journal's web page as Supplemantary Material Table 8), similarity ratios were found to between 0.087 and 0.593.The lowest (0.087) and highest (0.593) similarities were found between L11A and L33A, and between L25A and L29A, respectively.The minumum polymorphism was determined in L33A genotype with an average value of 0.2222 and the maximum polymorphism in L25A genotype with an average value of 0.404.The average value of genetic similarity according to IRAP markers in all specimens was calculated as 0.323.

Discussion
The average PIC value was calculated as 0.458 according to the ISSR assay.If a PIC value is >0.5, then polymorphism is at a high level, while if it is between 0.25 and 0.5, the polymorphism is normal.However, when the PIC value is 0.25, it is necessary to mention the weakness of the polymorphism (Botstein et al. 1980).The highest genetic similarity was found between the Cretan genotypes (LG2 and LG3), indicating that the biogeographic distribution is the factor for microsatellite polymorphism.It was remarkable that the similarity between Çeşme culture varieties (LCES1-2) and Cretan samples (LG1-2-3) was 0.629, indicating the closeness of the phylogenetic and genetic relations.The similarity value was pointed to the biogeographic source of the material used in culture studies.LCHIA (Chios variety) also showed the highest similarity (0.760) to L28B (C1, Kıyıkışlacık village / Milas).The highest ISSR polymorphism of the L28B (wild genotype) may be due to the change in primer binding sites, the independent alleles to occur and mutations in the microsatellite gene (Freudenreich et al. 1997).The phylogenetic tree constructed with the ISSR data shows that the specimens are separated according to their biogeographic positions.From the phylogenetical data, it was observed that the genotypes were divided into 12 groups and 9 of them were independent branches in the dendrogram.It can be explained by the emergence of isolated and differentiated genotypes that grow in various habitat conditions as adapted ecotypes.In the study of Zografou et al. (2010), UBC842, UBC850 and UBC856 primers were used, and the similarity ratios ranged from 68% to 12.5%.However, the similarity coefficient between the Chios varieties and the other specimens in our study was calculated between 0.760 and 0.093.This result indicates a wider distribution range compared to the results of Zografou et al. (2010).In another study with P. lentiscus (Abuduli et al. 2016), gene pool was investigated using SSR, RAPD and ISSR markers, and ITS regions in 24 different P. lentiscus wild genotypes, and the highest polymorphism was obtained by ISSR compared to other molecular markers (SSR, RAPD and ITS region) and the PIC value was calculated as 0.887.In the study using ISSR, the genotypes examined were divided into 5 subpopulations.As a result, the use of the ISSR markers was demonstrated to be a useful tool for genetic diversity analysis in wild genotypes of P. lentiscus and in future breeding studies (Abuduli et al. 2016).Considering the genetic similarity rates of the intraspecific variation obtained by the IRAP method, it was determined that LG1 (Cretan) was the closest specimen to the LCHIA (Chios variety.In terms of genetic similarity, the nearest sample to LCHIA genotype was L26A (C1, Güvercinlik village / Bodrum, 3m) and the furthest sample was L33A (C2, sea edge, Göcek / Fethiye, 62m), showing that species locations along seaside or in the lower parts of the mountains were not effective for genetic similarity.The dendrogram of IRAP data was examined in order to evaluate phylogenetic relations and it was shown that 3 main groups and 1 genotype (LBB) were separated, and retrotransposon mobilities resulting in high polymorphism were affected by the geographical locations of the plant samples.The results of Kılınç et al. (2014) (Kırdök & Çiftçi 2016).In addition, retrotransposon marker systems have also been used in taxonomic studies of the Persian species of the genus Pistacia (P.vera, P. khinjuk, P. atlantica) (Ghaemmaghami et al. 2013).In another study, IRAP analysis resulting in high polymorphism was performed on P. vera, P. khinjuk, P. eurycarpa and P. atlantica (Amirbakhtiar & Sorkheh 2015).Genetic diversity analysis using SCoT and IRAP markers in wild Pistacia species found an average PIC value of 0.32 for IRAP (Sorkheh et al. 2016).In conclusion, the applications of molecular markers are useful not only in the intraspecific variations, but also in interspecific analysis of Pistacia species.In a study using RAPD and AFLP markers, Pistacia was divided into two groups in a dendrogram segregated as evergreen P. lentiscus and the other group as deciduous (Golan-Goldhirsh et al. 2004).UPGMA analysis using AFLP markers showed clustering of P. terebinthus with P. mexicana and P. lentiscus implying their phylogenetic relations (Kafkas 2006).Golan-Goldhirsh et al. (2004) also reported that close genetic relationships in Pistacia species gave more reliable results in sex taxonomy compared to morphological classification.In the present study, the PCA analysis identifed different groups of 35 accessions in close proximity to each other.The results from PCA analysis based on ISSR and IRAP data showed that the variation in the P. lentiscus gene pool in Anatolia was high, genotypes were separated as groups, geographically isolated accessions were clearly separated from the others, and geographically related specimens were grouped together.On the other hand, significant differences were calculated among the accessions based on morphological parameters.Supportingly, phytochemical and morphological characters of P. lentiscus genotypes collected from Variables (axes F1 and F2: 48.71 %)

Fig. 5 .
Fig. 5. PCA output of the populations based on ISSR and IRAP data.

Table 1 .
The details of the localities of Pistacia lentiscus L. specimens included in the study.

Table 2 .
The list of the ISSR and IRAP primers with their corresponding Ta°C values and 5'-3' sequences.

Table 4 .
The correlation values between the character pairs.

Table 5 .
ISSR and IRAP analysis reaults for each primer.
and Koç et al. (2014) on P. lentiscus based on IRAP and AFLP techniques were supported with the findings of amplification data obtained in our present study in relation to the LTR2 region.Kırdök & Çiftçi (2016) used iPBS, REMAP and IRAP techniques of retrotransposon markers for genetic diversity analysis of Pistacia genus.They found high genetic similarity between P. vera -P.khinjuk, P. atlantica -P.mutica and P. terebinthus -P.palaestina and it was reported that the IRAP technique of retrotransposon-based DNA markers for 7 Mediterranean Pistacia species (total 35 samples) could be used in molecular characterization and molecular breeding trials