Mehrana Koohi Dehkordi·Tayebeh Beigzadeh·Karim Sorkheh

Abstract Because there are thousands of peach cultivars,cultivar classification is a critical step before starting a breeding project.Various molecular markers such as simple sequence repeats(SSRs)can be used.In this study,67 polymorphic primers produced 302 bands.Higher values for SI index(1.903)suggested higher genetic variability in the genotype under investigation. Mean values for observed alleles (Na), expected heterozygosity (He),effective alleles(Ne),Nei’s information index(h),and polymorphic information content(PIC)were 4.5,0.83,5.45,0.83,and 0.81,respectively.The dendrogram constructed based on Jaccard’s similarity coefficients outlined four distinct clusters in the entire germplasm.In addition,an analysis of molecular variance(AMOVA)showed that 70.68% of the total variation was due to within-population variation,while 29.32% was due to variation among populations. According to this research, all primers were successfully used for the peach accessions.The EST-SSR markers should be useful in peach breeding programs and other research.

Keywords Expressed sequenced tags(EST)·Simple sequence repeats(SSR)·Prunus persica L.·Genetic diversityl

Introduction

Peaches(Prunus persica L.)are grown all over the world in both subtropical and cold climates.However,its narrow genetic basis may limit selection efficiency in breeding programs(Cipriani et al.1999;Dirlewanger et al.2002).Self-fertility in peach is a critical tool that is used to authenticate the origin of selections and thus preclude mistakes.

Genetic variability in similar groups of peaches has been analyzed using simple sequence repeat(SSR)markers,or microsatellites because they are co-dominant,hypervariable,and highly informative(Blair et al.2010;Du et al.2012). Diverse markers, including morphological, biochemical,and molecular markers have also been used for various crop species (Rao 2004). Unlike DNA-based markers,however,morphological and biochemical markers can be influenced by environmental factors.PCR-based markers are also popular since they do not require predetermined sequence information.

A microsatellite represents a band of repetitive DNA,where specific DNA motifs are repeated,with their length ranging from 2 to 5 base pairs.Microsatellites occur in numerous regions within the genome of an organism,with great genetic diversity(Kantety et al.2002;Aranzana et al.2002,2003).Conversely,expressed sequence tags(ESTs)consist of single-pass DNA sequence classes,which offer direct information regarding expression of the gene.They serve as a source of SSRs(Martínez-Gómez et al.2003;Nagaraj et al.2007;Sorkheh et al.2016).Traditional techniques of developing genomic SSR markers involve the construction of genomic libraries and the isolation and sequencing of clones,which are time-consuming and laborintensive.Such techniques have been used in studies on Prunus species for peach(Sosinski et al.2000;Yamamoto et al.2002;Vendramin et al.2007),almond(Xu et al.2004;Xie et al.2006),apricot(Decroocq et al.2003;Hagen et al.2004),and mei(Li et al.2014;Wang et al.2014). Computational methods for isolating SSRs and developing SSR markers from EST-SSR are favored over conventional approaches.

The EST sequence database offers a platform for a straightforward analysis of EST sequences,facilitating the development of EST-SSR markers.EST sequences bear repetitive units,which makes them redundant(Chen et al.2006).Such bioinformatic methods have been used only for the peach genome among Prunus species(Chen et al.2014;Dettori et al.2015).

EST-SSR markers are popular for a number of reasons.Developing markers from a sequence in a database is rapid,relatively easy,and economical.Suitable programs can discover any SSR types,while enrichment cloning exclusively detects SSRs with predefined motifs.Owing to the preferential association of non-repetitive segments of plant genomes with SSRs,they represent a common EST component(Morgante et al.2002).In addition,EST-SSRs are physically associated with expressed genes,so that they represent functional markers that are of interest for markerassisted selection (MAS) (Andersen and Lubberstedt 2003).Furthermore,primer target sequences that are found in regions of expressed DNA ought to be well conserved,there by facilitating marker transferability across taxonomic borders(Varshney et al.2005).

The current study explores the molecular diversity of 293 peach accessions derived from seven states in Iran.We use a subset of the Iranian core collection and developed novel EST-SSR primers to detect genetic variation.The genotyping data were applied to comprehend relationships among the 293 accessions from the core set of peach germplasm under study,while isolating genetically diverse lines to use for peach genetic enhancement.

Materials and methods

Plant material and total genomic DNA isolation

We sampled 293 genotypes from a subset of the core collection of peach germplasm at the Institute of Fruit Research, Shahrekord; Table 1 presents details of the peach accessions.

Total genomic DNA was extracted using the method of Murray and Thompson(1980)as modified and adapted by Weising et al.(1995)and detailed by Sorkheh et al.(2007).

Preparing of EST sequences and assembly

All sequences were downloaded from GenBank(ftp://ncbi.nlm.nih.gov/genbank/genomes/) (Table S1).To develop longer and fewer excess sequences,freely accessible ESTs were collected from CAP3 (Huang and Madan 1999;Whitfield et al.2001;Pertea et al.2003).

Identification of SSR and characterization

The aim was to eliminate redundancy in EST sequences to obtain contiguous sequence(contigs)that can be used for SSR analysis.For the aim of SSR identification,CAP3 contig and singleton preparation were confederated to form non-excess sequence data.Genomic SSRs were identified using GMATo (http://sourceforge.net/p/GMATo). The identified SSRs were characterized as described by Sorkheh et al.(2016).

EST-SSR primer prescreening and selection

Of 175 EST-SSR primers screened,only 67 primers yielded clear and reproducible polymorphic bands and were thus selected to analyze genetic diversity.Table S2 presents the primer sequence and the annealing temperatures of the 67 selected.

Table 1 Collection information for the 293 peach accessions studied

PCR amplification

PCR reactions were conducted in 25 μL reactions,according to the protocol presented by Sorkheh et al.(2016).PCR products were separated by electrophoresis in 3% agarose(SinaClone Ltd,Iran)and stained with Safe stain(SinaClone,Karaj,Iran).The molecular size standard was Invitrogen1 Kb Plus DNA Ladder(Thermo Fisher Scientific,Waltham,MA,USA).Most of the band patterns presenting high-quality polymorphism and critical information had a background smear.To minimize the smear,we added formamide(2% ,v/v)to the mixture.Generated profiles were reproduced at least three times.

Analysis of genetic diversity

The amplified bands were sorted into the present(1)or absent(0)groups for every marker.Binary matrices were created based on the marker data and then used to derive Jaccard’s similarity coefficients(Jaccard 1908)as follows:

where niis the number of bands present in cultivar I but absent in cultivar j,njis the number of bands present in cultivar j and absent in cultivar i,and nijrepresents the number of bands present in both cultivars.

A cluster analysis was used to construct a dendrogram using the similarity matrix data and NTSYS-pc 2.10 software(Rohlf 2002).

For quantitatively assessing the number of groups in the panel,a Bayesian analysis was carried out in STRUCTURE v 2.3(Pritchard et al.2000),assuming that there is an admixture of independent allele frequencies.Analyses were also carried out for the number of subgroups(ΔK)present,with K as the number of clusters or subpopulations,considering change rate in the data’s log probability(Evanno et al.2005),which ranges from 1 to 10.Ten independent runs were carried out per K value,beginning with 10,000 burn-in periods and 100,000 Markov chain Monte Carlo(MCMC)replications.A 20,000 burn-in and 250,000 MCMC replications were selected as the appropriate fit for the data at K=4.We calculated values for polymorphism information content (PIC)to investigate diversity levels with EST-SSR markers in line with Anderson et al.(1993),based on the following formula:

where pijdenotes jth allele frequency in the ith EST-SSR locus and g denote number of locus.

Binary data was also subjected to a principal component analysis(PCA)to investigate the structure of the 293 peach accessions using the EIGEN module of NTSYSpc 2.10(Rohlf 2002).The genetic variability features of each geographic group,including the proportion of polymorphic loci,number of alleles per loci,effective allele quantity per locus,Nei’s gene diversity,as well as SI index,were computed in POPGENE version 3.2(Yeh et al.2000).

AMOVA was carried out,after establishing optimum population numbers,in Arlequin version 3.1(Excoffier et al.2005)to evaluate genetic variation both between and within populations.

Results

Variation of allelic among accessions from ESRSSRs

The 67 EST-SSRs used to determine genetic variability in 293 P.persica accessions yielded amplicons sites ranged from 100 to 650 bp.The 67 EST-SSRs were polymorphic(100% )and gave rise to 302 alleles.The number of alleles for each EST-SSR varied between 2 (ESTSSR59,ESTSSR58,ESTSSR48,ESTSSR47,ESTSSR32,ESTSSR31,ESTSSR13,and ESTSSR12)and 12(ESTSSR20),for a mean of 4.5.The PIC values for the EST-SSR primers evaluated were fairly high,ranging from 0.68 to 0.86,for a mean of 0.81.This value implies primer effectiveness,polymorphism,and validity(Table 2).

The highest effective number of alleles(6.929)was found for ESTSSR65,ESTSSR61,ESTSSR58,ESTSSR46,ESTSSR40,ESTSSR33,ESTSSR30,ESTSSR20,ESTSSR18,and ESTSSR10 and the fewest (3.493) for ESTSSR66,ESTSSR62,ESTSSR59,ESTSSR50,ESTSSR47,ESTSSR41,ESTSSR34,ESTSSR31,ESTSSR22,ESTSSR19,ESTSSR11,and ESTSSR3(Table 2).In addition,the highest levels of expected heterozygosity(He)were estimated for ESTSSR10,ESTSSR21, ESTSSR30, ESTSSR41, ESTSSR52, and ESTSSR63(0.885)and the lowest for ESTSSR3,ESTSSR12,ESTSSR23,ESTSSR32,ESTSSR43,ESTSSR45,ESTSSR54,and ESTSSR65(0.725),for a mean of 0.83(Table 2).

The Nei’s value for diversity of gene ranged from 0.7279 to 0.8862,with the highest values for ESTSSR10,ESTSSR27,ESTSSR39,ESTSSR48,and ESTSSR63 and lowest for ESTSSR3,ESTSSR11,ESTSSR28,ESTSSR40,ESTSSR49,ESTSSR54,and ESTSSR64(mean 0.8351).In addition,the SI index ranged from 1.4457 to 2.1327(mean 1.903).The highest SI indices were found for ESTSSR63,ESTSSR52, ESTSSR44, ESTSSR31, ESTSSR23, and ESTSSR10, the lowest for ESTSSR64, ESTSSR53,ESTSSR45, ESTSSR32, ESTSSR24, ESTSSR11, and ESTSSR3(Table 2).

Na,observed number of alleles;Ne,effective number of alleles; He, expected heterozygosity; Ho, observed heterozygosity;polymorphic information content,PIC;h,Nei’s index;SI,Shannon’s information index.

Genetic association among the peach accessions

Genetic relationships among 293 peach accessions were analyzed using the UPGMA clustering technique,as shown in Fig.1.To explore further the genetic association among the studied germplasm,a cut-offpoint(H=0.74)was defined using the technique of Sorkheh et al.(2007)for clustering accessions.The dendrogram split seven populations(293 accessions)into four distinct groups.Group I consisted of LORD_1, LORD_4, LORD_DORF2,LORD_2,LORD_DOR,LORD,and LORD_12.Group II consisted of BEN_HOR4, BEN_F2, ARDL_BH, and ARDL. Group III consisted of SK_F1, SK, SK_F2,SK_KC,SMN_F1,BEN_F1,and LORD_3,while group IV consisted of BEN, BEN_HOR2, BEN_HOR, and BN_F3.Nei’s genetic distance between pairwise population matrixes is illustrated in Table 3.

A Bayesian analysis in STRUCTURE software confirmed the clusters obtained from the UPGMA dendrogram and PCA,based on the molecular data(Fig.2).The most probable value of K from the Bayesian analysis was 4.This result implies that variation was divided into four clusters,so that there were four major clusters among the peach populations under study.This also confirmed the clusters revealed by the UPGMA and PCA.

Principal component analysis(PCA)

The cluster patterns observed were further corroborated by a PCA(Table 4).The PCA results derived from the ESTSSR markers suggested that 67.07% of the total variance observed could be attributed to three first principal components.PC1,PC2,and PC3 accounted for 48.93,9.87,and 8.27% of the variation,respectively.

Analysis of molecular variance(AMOVA)

According to the results of AMOVA,the variability among the populations under study was low,at 29.32% .However,variation within the populations represented 70.68% of the total variation in the studied germplasm(Table 5).

Table 2 Pairwise population matrix of Nei’s genetic distance among seven populations using new EST-SSR markers

Table 2 continued

Discussion

SSR markers were detectable in both transcribed and the non-transcribed sequences,and these were referred to as EST-SSRs and genomic SSRs,respectively.A major merit of EST-SSRs over genomic SSRs is their high level of transferability. That is, EST-SSRs may be transferred across species that are closely related(Guichoux et al.2011).Consequently,in this study,transcriptome ESTSSRs and bioinformatics(Vendramin et al.2007;Ding et al.2017)were used to detect genetic variation among the selected P.persica L.accessions.

Distinguishing different accessions within species or clones is a critical step for breeding programs,which begin with the investigation and detection of variation among genotypes.Investigating genetic variability is also a key aspect of applied plant improvement,since it facilitates the selection of optimal parents.In addition,effective assessment of a germplasm offers a reliable and scientific basis for selecting appropriate parents or donors to participate in fruit breeding program,and to breed for particular environments(Kumar et al.2012).

Consistent and effective use of molecular markers,for example,EST-SSRs,in the survey of variation of genetic tree crops demands selecting and applying primers that will offer apparent, reliable, definite, as well as adequate information necessary to investigate variation and potential of peach germplasm(Testolin et al.2000;Arolu et al.2012).In the present study,the number of polymorphic loci varied depending on the primer.The EST-SSR markers used in the present study amplified distinct bands,revealing clear patterns among the 293 P.persica germplasm.

The 67 EST-SSR markers revealed 302 alleles,with an average of 4.5alleles per marker.This value is relatively high compared to those established for other species in Prunus genus,including almond(P.communis Fritsch),with an average of 6.64(Shiran et al.2007),implying higher discrimination power;however,the value is consistent with those obtained for peach in other studies.For instance,a study on 50 peach cultivars reported an average of 4.5 alleles per locus(Testolin et al.2000)while another study applying 16 polymorphic SSR markers reported an average of 7.3 alleles per locus(Aranzana et al.2002).It is evident that the number of alleles(Na)is related to the sample size and the number of SSRs applied.

PIC values ranged from 0.68 to 0.863,with an average of 0.81.The results of the present study are consistent with those obtained by Hu et al.(2011),who estimated the genetic diversity of a Cucumis sativus L.germplasm using SSR and EST-SSR markers. In a study on common almond,Ma et al.(2003)used already developed ESTSSRs and reported low levels of polymorphism in almonds and high levels of polymorphism in other Prunus species.The results of the study above suggested that EST-SSR markers obtained from a single Prunus species could be applied to other related Prunus species.In addition,in thepresent study,allele number ranged from 2 to 12,with an average of 4.5(Table 2).This outcome is similar to that obtained in studies exploring genetic diversity in watermelon populations using EST-SSR markers(Hu et al.2011;Mujaju et al.2013).

Table 3 Pairwise population matrix of Nei’s genetic distance among seven populations(Pop.)using new EST-SSR markers

Fig.2 Grouping of 293 peach accessions genotyped at 67 novel EST-SSR loci based on STRUCTURE analysis(top)and the log likelihood for each K,Ln P(D)=LogK probability(bottom)

Table 4 Principal component analysis and percentage variation in first three principal components for 293 peach accessions by EST-SSR markers

Table 5 AMOVA within and among seven studied populations from a subset of the core peach germplasm using new EST-SSR markers

Our EST-SSR markers further demonstrated varying levels of diversity among the P.persica accessions.The SI index ranged from 1.44 to 2.13,with an average of 1.90.Fathi et al.(2008),in a study of various almond cultivars,reported SI indices from 0.35 to 2.6(average 1.79).Higher values for the SI index in the present study also suggest higher variability among the genotypes under investigation.The Nei’s gene diversity value ranged from 0.72 to 0.88(mean 0.83)(Table 2).In another study using nuclear SSR markers from wild almonds,however,gene diversity ranged from 0.73 to 0.93(Zeinalabedini et al.2007).

In the present study,expected heterozygosity ranged from 0.72 to 0.88, with a mean of 0.83. Expected heterozygosity was higher and lower in other species within the genus,including in apricot,P.armeniaca L.with a mean of 0.517(Zhebentyayeva et al.2003),and in cherry,with a mean of 0.946(Cantini et al.2001).Generally,allele richness and‘Ho’in the peach accessions under study from southern Iran were poor when compared to those of other fruit trees(Lamboy and Alpha 1998;Benson et al.2001;Zhen et al.2004).Nevertheless,the allele frequency and He and Ho were in agreement with results of previous studies(Wang et al.2002).The variation observed may be attributed to the effects of the different breeding systems in various Prunus species,as well as differences in ploidy,on overall genomic diversity,considering P.persica is principally self-fertilizing and diploid.When comparing P.persica populations,however,the mean value of 0.83 was greater than 0.35 reported by Aranzana et al.(2002)and 0.32 by Cipriani et al.(1999).The superior level of He and Ho in the present study could be due to accessions that varied in characteristics such as color,flavor,texture,and ripening date.

Observed heterozygosity ranged between 0.088 and 0.821,with a mean of 0.368.Amirbakhtiar et al.(2006)reported values ranging from 0.1 and 0.87,with a mean of 0.5,while Martínez-Gómez et al.(2003)reported higher levels of heterozygosity(0.38 to 0.88)in almond cultivars than in peach cultivars(＞0.36).Another study revealed that observed heterozygosity per locus was 0.696 in Chinese almond cultivars and 0.583 in international cultivars(Xie et al.2006).

Parameters used to assess genetic diversity such as expected heterozygosity,Nei’s gene diversity,SI index,and PICs are positively correlated with the number of alleles.In the present study,EST-SSR markers revealed the highest values of expected heterozygosity(0.83),Nei’s gene diversity(0.83),SI index(1.90),and PIC(0.81)(Table 2).Considering that higher values in the parameters indicate higher diversity, the EST-SSR markers were considered more appropriate for a genetic analysis of the P.persica accessions.

A cluster analysis offered a better constancy of the relationships among the 293 peach accessions using ESTSSR.The EST-SSR markers revealed eight major clusters among the 293 accessions from different geographical populations(Fig.1).The EST-SSR markers could have produced eight clusters because of the mode of action of EST-SSR markers,which target larger numbers of specific repeated sequences within the centromere regions,which can greatly permeate clustering(Parsons et al.1997).Such clustering is evidence of the appropriateness and compatibility of EST-SSR markers in genetic variation analysis of peach germplasm.

The PCA revealed 67.07% total variation among all accessions.In addition,PC1,PC2,and PC3 accounted for 48.93,9.87,and 8.27% ,respectively,from the Eigen vectors (Table 3). According to Fu et al. (2014), cluster analysis and PCA results for genetic variability were close to those for celery(Apium graveolens L.)cultivars when EST-SSR markers were applied.Similar results were found in a study analyzing genetic diversity in soybean using EST-SSR markers(Zhang et al.2013).

The results of AMOVA showed that a higher proportion of the variation was due to variation within populations(70.68% )than among populations(29.32% )in the P.persica populations under study(Table 5).According to Li et al.(2011),evaluation of the genetic diversity within a population is critical for characterizing germplasm and offers essential insights that can facilitate understanding of evolutionary trajectories, conservation, exploitation, as well as the establishment of breeding programs.

Conclusions

In the present study,high levels of genetic diversity in peach germplasm show the effectiveness of EST-SSR markers in identifying polymorphisms among the studied peach cultivars.The present results represent the first report that explores genetic diversity in a P.persica germplasm using EST-SSR markers.For broadening the genetic base or enhancing P.persica populations,populations with the lowest genetic similarity could be recruited as parents;using two genetically distant populations for breeding will result in greater genetic diversity and enhance productivity indices for yield and general quality.