Heredity (2004) 93, 631–638 & 2004 Nature Publishing Group All rights reserved 0018-067X/04 $30.00 www.nature.com/hdy Genetic relationships between diploid and allotetraploid cherry species (Prunus avium, Prunus  gondouinii and Prunus cerasus)

M Tavaud1,3, A Zanetto1, JL David2, F Laigret1 and E Dirlewanger1 1INRA, UREFV, BP81, 71, Avenue E Bourleaux, 33883 Villenave d’Ornon, France; 2INRA, UMR DGPC, Domaine de Melgueil, 34130 Mauguio, France

Prunus avium L. (diploid, AA, 2n ¼ 2x ¼ 16), Prunus cerasus correspondence analysis allowed us to clearly distinguish L. (allotetraploı¨d, AAFF, 2n ¼ 4x ¼ 32) species, and their each species, to identify misclassifications and to assign hybrid Prunus  gondouinii Rehd., constitute the most widely unknown genotypes to one species. We showed that there cultivated cherry tree species. P. cerasus is supposed to be are specific alleles in P. cerasus, which are not present in the an hybrid species produced by the union of unreduced A genome of P. avium and which probably come from the F P. avium gametes and normal P. fruticosa gametes. A genome of P. cerasus. The frequencies of each marker in the continuum of morphological traits between these three A and the F genomes were estimated in order to identify A species makes their assignation difficult. The aim of this and F specific markers. We discuss the utility of these paper is to study the genetic relationships between tetraploid specific markers for finding the origin of the A and F genomes and diploid cherry species. In all, 114 genotypes belonging to in the allopolyploid species. these species were analyzed using 75 AFLP markers. Heredity (2004) 93, 631–638. doi:10.1038/sj.hdy.6800589 The coordinates of these genotypes on the first axis of a Published online 8 September 2004

Keywords: Prunus avium; Prunus cerasus; AFLP markers; species identification; genetic diversity

Introduction tion between P. avium (producing unreduced gametes) and P. fruticosa (Figure 1). This origin was first suggested The most widely cultivated cherry trees belong to Prunus by Olden and Nybom (1968), who observed that artificial avium L. and P. cerasus L. species. Along with P. fruticosa hybrids between tetraploid P. avium and P. fruticosa were Pall., these species and their interspecific hybrids very similar to P. cerasus. Isozyme analyses, genomic constitute the Eucerasus section of the Cerasus subgenus, in situ hybridization and karyotype analysis further based on morphological criteria (Rehder, 1947; Kruss- confirmed the hybrid origin of P. cerasus (Hancock and mann, 1978). This classification and the monophyletic Iezzoni, 1988; Santi and Lemoine, 1990a; Schuster and origin of the Eucerasus clade have been confirmed by Schreiber, 2000). The patterns of inheritance of seven chloroplastDNA variation analysis (Badenes and Parfitt, isozymes in different crosses of sour cherry indicated 1995). that P. cerasus may be a segmental allopolyploid (Beaver P. avium includes sweet cherries, cultivated for human and Iezzoni, 1993). Recent studies based on cpDNA consumption and wild cherry trees grown for their wood markers detected two distinct chlorotypes in P. cerasus, (Webster, 1996). This species is diploid (AA, 2n ¼ 2x ¼ 16) which strongly suggest that crosses between P. avium and and its natural range covers the temperate regions P. fruticosa have occurred at least twice to produce sour of Europe, from the Northern part of Spain to the cherry (Badenes and Parfitt, 1995; Iezzoni and Hancock, Southeastern part of Russia (Hedrick, 1915). P. fruticosa, 1996; Brettin et al, 2000). Moreover, these results suggest the ground cherry tree, is a tetraploid wild species that, most of the time, P. fruticosa was the female (2n ¼ 4x ¼ 32) thought to be autotetraploid (FFFF), some- progenitor of P. cerasus, but in few cases P. avium was times used as rootstocks for other Prunus species. This the female parent, via the formation of unreduced species is widespread over the major part of central ovules. Triploid hybrids through the fusion of normal Europe, Siberia and Northern Asia (Hedrick, 1915). P. gametes of P. avium and P. fruticosa occur naturally but cerasus, the sour cherry tree, is cultivated for fruit, used in remain sterile. Moreover, they are not clonally propa- jam or liquor. It is an allotetraploid species (AAFF, gated by man since they exhibit many unfavorable P. 2n ¼ 4x ¼ 32), thought to result from natural hybridiza- fruticosa traits (Olden and Nybom, 1968). The duke cherries, which result from crosses between Correspondence: Dr M Tavaud, INRA, UREFV, BP81, 71, Avenue E P. avium and P. cerasus, are cultivated at a much smaller Bourleaux, 33883 Villenave d’Ornon, France. scale. Different names have been given to this species E-mail: [email protected] including P. acida Dum, Cerasus regalis, P. avium ssp 3Current address: INRA, UMR DGPC, Domaine de Melgueil, 34130 Mauguio, France. regalis, but the name used currently is P. Â gondouinii Received 27 November 2003; accepted 16 July 2004; published Rehd. (Faust and Suranyi, 1997; Saunier and Claverie, online 8 September 2004 2001). This species is allotetraploid (AAAF, 2n ¼ 4x ¼ 32), Genetic relationships between three cherry species M Tavaud et al 632 cerasus and to identify the origin of the A genome in both Prunus fruticosa Prunus avium tetraploid species (P. cerasus and P. Â gondouinii). 2n=4x=32 2n=2x=16 FFFF AA * Materials and methods

Plant material Prunus cerasus In all, 114 genotypes were studied: 68 P. avium genotypes 2n=4x=32 * (33 cultivated and 35 wild) sampled throughout the AAFF natural range of the species, 25 P. cerasus genotypes and 9 P. Â gondouinii genotypes according to morphological characters (Table 1). In total, 12 French genotypes which had not been previously studied were not assigned to a species. The 33 cultivated sweet cherry trees were old Prunus x gondouinii cultivars, from 13 different countries. The 35 wild cherry =Prunus acida Dum trees were from several European forests. Sour and duke 2n=4x=32 cherry trees and unassigned genotypes were from the AAAF French germplasm, mainly maintained in INRA orchards, at the ‘Unite´ Expe´rimentale d’Arboriculture de Bordeaux’. Figure 1 Hypothesis for the relationships between the four species belonging to the Eucerasus section. *P. avium is thought to produce diploid gametes. A and F are haploid genomes coming from P. AFLP analyses avium and P. fruticosa respectively. Genomic DNA was extracted from young leaves, follow- ingthemethoddescribedbyViruelet al (1995). We stemming from the fertilization of sour cherry by essentially followed the AFLP protocol developed by Vos unreduced gametes of sweet cherry (Iezzoni et al, 1990). et al (1995) with minor modifications. DNA was digested These hybrids are often sterile, due to disturbances with EcoRI and MseI restriction enzymes. Digestion was during meiosis, but they are clonally propagated by man. performed for 4 h at 371C in a final volume of 17.5 ml The duke cherry trees present tree and fruit character- containing 10 mM Tris-Ac, 10 mM MgAc, 50 mM KAc, istics intermediate between their progenitors. 5mM DTT, 25U EcoRI (Pharmacia), 4 U MseI(New Great morphological variation exists among P. avium, England Biolabs) and 250 ng of genomic DNA. Two P. fruticosa and P. cerasus species. Multivariate analysis on linkers, one for the EcoRI sticky ends and the other for sour cherry revealed continuous variation between the the MseI sticky ends, were ligated for 3 h at 371Ctothe P. avium, P. cerasus and P. fruticosa traits throughout the digested DNA by adding 2.5 ml of a mix containing geographic distribution of the species. In Western 2.5 pmol EcoRI linker, 25 pmol MseI linker, 8 mM ATP, Europe, P. cerasus trees look like P. avium, whereas in 10 mM Tris-HAc, 10 mM MgAc, 50 mM KAc, 5 mM DTT Eastern Europe P. cerasus are closer to P. fruticosa (Hillig and 0.85 U T4 DNA ligase (Pharmacia). This ligation and Iezzoni, 1988; Krahl et al, 1991). This continuum of product was diluted five-fold. A first preselective PCR morphological characteristics makes species assignation amplification was performed using EcoRI þ AandMseI þ difficult when considering only phenotypic traits. C primers in a 50 ml mix of 20 mM Tris–HCl (pH 8.4), Genetic resource conservation, essential for future 50 mM KCl, 1.5 mM MgCl2, 0.2 mM of each dNTP, 30 ng of breeding programs, requires a good characterization of each primer, 1 U Taq DNA polymerase (GibcoBRL) and 5 ml germplasm genetic diversity and a proper assignation of of diluted ligation product. The reaction was carried out in individual genotypes to species. Molecular markers a Perkin-Elmer 9600 thermocycler and the samples were supply tools to analyze genetic diversity in Prunus subjected to 28 amplification cycles with three steps of 30 s species (Baird et al, 1996). They could be helpful by at 941C, 60 s at 561C and 60 s at 721C. The preamplification giving an accurate and unambiguous assignation of each products were diluted 10-fold and used as starting genotype to a particular species. Isozymes have been material for the selective radioactive amplification. already used to characterize genetic diversity in P. avium For selective amplification, EcoRI and MseI primers and P. cerasus, but the number of available isozymes is with three and two additional nucleotides, respectively, limited and their polymorphisms are low (Santi and were used (Table 2), the first one being 33P-labelled using Lemoine, 1990b). AFLP markers could be useful because T4 polynucleotide kinase. The PCR reaction was per- they are reproducible, they need no previous information formed in a 20 ml volume of 20 mM Tris–HCl (pH 8.4), 33 about the genome and they allow the detection of many 50 mM KCl, 1.5 mM MgCl2, 5 ng [ P] EcoRI primer, 30 ng markers in few analysis (Vos et al, 1995; Jones et al, 1997). MseI primer, 1 U Taq DNA polymerase (GibcoBRL) and They are well suited for studying unexplored species and 5 ml of diluted preamplified DNA. The selective ampli- have already been used successfully to discriminate fication was carried out using the following cycling different species with varying ploidy levels in Gossypium parameters: 11 cycles of 30 s at 941C, 30 s at 651C, 60 s at and Oryza (Aggarwal et al, 1999; Abdalla et al, 2001). 721C, in which the annealing temperature was lowered The aim of this work was first to determine whether by 0.71C per cycle, followed by 24 cycles of 30 s at 941C, AFLP markers could be useful in assigning each cherry 30 s at 561C and 60 s at 721C. genotype accurately to one species (P. avium, P. cerasus or The PCR products, in which an equal volume of load P. Â gondouinii). We also wanted to find specific markers buffer (98% formamide, 10 mM EDTA, 0.05% bromo- of the different genomes (A and F) of these cherry phenol blue and 0.05% xylene cyanol) had been added, species. These markers would be useful for comparing were denatured 5 min at 941C and immediately placed on the diversity of the A genome between P. avium and P. ice. In all, 2 ml of each sample was loaded on a

Heredity Genetic relationships between three cherry species M Tavaud et al 633 Table 1 (a) A priori species assignation, geographical origin and sample number of the cultivated and wild genotypes studied. (b) Names of cherry varieties analysed and their species assignation according to their coordinates on the first axis of the factorial analysis (a)a

P. avium priori assignations Geographical origin Sample number P. avium posteriori assignations

Cultivated cherry Wild cherry P. av P. gon P. cer

P. avium Czech republic 3 3 Canada 1 1 France 10 8 14 4 Germany 3 5 8 Georgia 4 4 Great Britain 2 2 4 Greece 2 2 Italy 3 5 5 3 Japan 1 1 Poland 1 1 Romania 2 3 5 Slovakia 1 1 Spain 5 3 2 Sweden 2 2 Switzerland 2 2 USA 1 1 Yugoslavia 2 1 1 Total 33 35 58 7 3 P. cerasus France 25 2 2 21 P. Â goundouinii France 9234 Undetermined genotypes France 12 2 4 6 Total 114

(b)

Variety names Origin a priori assignation a posteriori assignation

Bigarreau Pelissier France P. avium P. avium Guigne noire Luisante France P. avium P. avium Cœur de Bœuf France P. avium P. avium Koromilokeraso Vitalou Greece P. avium P. avium Petrokeraso Tragano Achaias Greece P. avium P. avium Pivovka Czech republic P. avium P. avium Chlumecka cerna Czech republic P. avium P. avium Sakvicka Czech republic P. avium P. avium Bigarreau Cœur de Noyon France P. avium P. avium Knauffs Schwarze Germany P. avium P. avium Karina Germany P. avium P. avium Bu¨ ttners Rote Knorpelkirsche Germany P. avium P. avium Bedford Prolific P. avium Great Britain P. avium P. avium Nutberry Black Great Britain P. avium P. avium Camus France P. avium P. avium Vittoria Italy P. avium P. avium Maru Rubys USA P. avium P. avium Van 2D 9–11 Canada P. avium P. avium Amar 153 Yugoslavia P. avium P. avium Bianca di Verona Italy P. avium P. avium Sato Nishiki Japan P. avium P. avium Arodel France P. avium P. avium Kunzego Poland P. avium P. avium Silva Romania P. avium P. avium Roz Amar de Ma`rculesti (RAM) Romania P. avium P. avium Ravenna Italia P. avium P. avium Fru¨he Luxburger Switzerland P. avium P. avium Webers Sa¨mpling Switzerland P. avium P. avium Marasca Yugoslavia P. avium P. cerasus Gros Guin de Cœur France P. avium P. Â gondouinii Guigne Boissie`re France P. avium P. Â gondouinii Guin des Charentes France P. avium P. Â gondouinii Meynard France P. avium P. Â gondouinii Frumi Canada P. cerasus P. avium Griotte jaune d’Ollins France P. cerasus P. avium Ferracida France P. cerasus P. cerasus Montmorency1 France P. cerasus P. cerasus Montmorency2 France P. cerasus P. cerasus Griotte de Moissac France P. cerasus P. cerasus Griotte Layat France P. cerasus P. cerasus

Heredity Genetic relationships between three cherry species M Tavaud et al 634 Table 1 (continued) Variety names Origin a priori assignation a posteriori assignation Griotte du Nord France P. cerasus P. cerasus Montmorency3 France P. cerasus P. cerasus Acide Haut Rhin1 France P. cerasus P. cerasus Acide Haut Rhin2 France P. cerasus P. cerasus Acide Haut Rhin3 France P. cerasus P. cerasus Cerise de Toussaint France P. cerasus P. cerasus Grinque Montmorency France P. cerasus P. cerasus Griotte pre´coce France P. cerasus P. cerasus Montmorency pleureur de Sauvigny France P. cerasus P. cerasus Iwa Poland P. cerasus P. cerasus Sabina Poland P. cerasus P. cerasus Lucyna Poland P. cerasus P. cerasus Cerise acide Mimbielle France P. cerasus P. cerasus Griotte de Moissac France P. cerasus P. cerasus Montmorency4 France P. cerasus P. cerasus Griotte du Lyonnais France P. cerasus P. cerasus Griotte de Provence France P. cerasus P. Â gondouinii Guigne Boissie`re France P. cerasus P. Â gondouinii Anglaise Hative1 France P. Â gondouinii P. avium Anglaise Hative2 France P. Â gondouinii P. avium Cerise Olivet Hative France P. Â gondouinii P. cerasus Cerise Olivet Tardive France P. Â gondouinii P. cerasus Cerise Reine Hortense France P. Â gondouinii P. cerasus Belle Magnifique France P. Â gondouinii P. cerasus Cerise Impe´ratrice Euge´nie France P. Â gondouinii P. Â gondouinii Gros guin noir de Gironde France P. Â gondouinii P. Â gondouinii Maynard France P. Â gondouinii P. Â gondouinii La Carre´e France unknown P. avium Petite joue vermeille France unknown P. avium Belle de Varennes France unknown P. cerasus Grosse cerise tardive France unknown P. cerasus Grosse griotte France unknown P. cerasus Triaux de Fondettes France unknown P. cerasus Belle Magnifique France unknown P. cerasus Cerise Royale Tardive France unknown P. cerasus Cerise Cure1 France unknown P. Â gondouinii Cerise Cure2 France unknown P. Â gondouinii Guin des Charentes France unknown P. Â gondouinii aA posteriori species repartition of genotypes in the three species was determined according to their coordinates on the first axis of the factorial analysis. For each species, total numbers are indicated in bold.

polyacrylamide gel (4.5% acrylamide/bisacrylamide 20:1, Table 2 Sequences of the primers used for AFLP selective 7.5 M Urea and 0.5  TBE) and were run at 95 W for 2 h. amplification and number of reducible polymorphic bands analysed After electrophoresis, gels were dried on a standard slab EcoRI/MseI Primer sequences Number of gel drier for 2 h and exposed for 5 days to an X-ray film. primers markers Each analysis was performed twice for each individual. Only intense, polymorphic and reproducible bands EcoRI+AGG 50-GACTGCGTACCAATTCAGG-30 20 0 0 were taken into account to generate a binary matrix 0–1, /MseI+CA 5 -GATGAGTCCTGAGTAACA-3 EcoRI+AGG 50-GACTGCGTACCAATTCAGG-30 17 where 1 and 0 denote, respectively, the presence and the /MseI+CG 50-GATGAGTCCTGAGTAACG-30 absence of a band. EcoRI+AAC 50-GACTGCGTACCAATTCAAC-30 22 /MseI+CG 50-GATGAGTCCTGAGTAACG-30 Statistical analyses EcoRI+AGA 50-GACTGCGTACCAATTCAGA-30 16 0 0 Species assignation: To properly assign based geno- /MseI+CA 5 -GATGAGTCCTGAGTAACA-3 types to their species, a correspondence analysis based Total 75 on the AFLP markers was first performed on the whole sample using the procedure CORRESP of the SAS software (SAS Intitute, 1985). An a posteriori species group was assigned to each genotype according to its considered as being at Hardy–Weinberg equilibrium coordinates on the first axis of this analysis. (random mating, nonoverlapping generations, large population size, mutation ignored and migration Variation of individual band number (IBN): As AFLP overlooked), this allelic frequency qi can be estimated 2 4 markers are dominant, the phenotype [1] corresponds by ^qi ¼½0Š in a diploid species and by ^qi ¼½0Š in an either to the genotype (11) or to the genotypes (01) or autotetraploid species, where [0] is the frequency of the (10), whereas phenotype [0] is only due to the (00) phenotype [0] for the locus i. Allele frequencies were genotype. Let qi be the frequency of allele 0 at locus i in calculated following this method for each locus in one species. If we assume that this species can be diploid P. avium species.

Heredity Genetic relationships between three cherry species M Tavaud et al 635 We wanted to know whether an increase of IBN in diversity of the A genome is lower in P. cerasus than in P. tetraploid P. cerasus species (AAFF), in comparison to avium. To estimate genetic diversity in the two species, P. avium, a diploid species (AA), was simply due to we used average heterozygosity (He), also known as polyploidization or could be attributed to specific gene diversity (Nei, 1978). Using only A genome-specific markers belonging to the F genome. Indeed, autopoly- markers, we computed He as Hei ¼ 2qi(1Àqi), where qi is ploidization increases IBN since the frequency of the the frequency of allele 0 at locus i in the species studied. 4 phenotype [1] is 1Àqi in a putative autotetraploid In P. avium, such frequency is qAi , whereas in pP.ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi cerasus, 2 species, whereas this frequency is 1–qi in diploid species this frequency is calculated by qAi cerasus ¼ ½0Šcerasus for each locus. IBN was calculated as the sum of all using only A-specific markers. For each species, we amplified bands [1] in each genotype considering all the calculated He as the mean of Hei. We computed AFLP markers. To determine the effect of autopolyploi- Hei aviumÀHei cerasus for each locus and tested the mean dization on the IBN, we simulated 500 autotetraploid difference using a sign test. genotypes (AAAA), formed by the autotetraploidization of P. avium with identical allelic frequencies qi to P. avium. For each simulated genotype, the phenotype at each Results locus i was randomly assigned to [0] or [1] with 4 4 probabilities ^qi and 1 À ^qi , respectively, where ^qi is the Species assignation estimated allele 0 frequency in P. avium. TheP mean value A total of 75 polymorphic AFLP markers were detected i¼75 2 of the IBN, in each species, isP computed as 1 ð1 À ^qi Þ using the four primer pairs, as indicated in Table 2. On i¼75 4 in diploid species, and 1 ð1 À ^qi Þ in tetraploid the two first axes of the correspondence analysis based species, where i denotes the locus, ^qi the allele 0 on the 75 AFLP markers, three clear-cut groups appeared frequency in the species. We compared the IBN (Figure 2). The first axis corresponds to 31.5% of total distribution in P. avium, the ‘simulated’ autotetraploid inertia. The P. cerasus and P. avium genotypes mainly species, P. cerasus and P.  gondouinii. compose the two extreme groups with some misclassi- fied individuals. In the P. avium group, one Canadian Determination of genome-specific markers: To compare variety and one old French variety were a priori declared genetic diversity in each genome, among species with as belonging to P. cerasus; two old French varieties were different ploidy levels, the determination of genome- previously classified as P.  gondouinii (Table 1). The P. specific markers is needed. In an allotetraploid species cerasus group contains three and four genotypes pre- like P. cerasus (AAFF), markers could be present viously classified as P. avium and P.  gondouinii, respec- specifically in one genome or in both genomes. We tively. The intermediate group contains nine cannot distinguish if the phenotype [1] in P. cerasus is due misidentifications (seven presumed P. avium and two P. to marker amplification in A or/and F genome. We cerasus), four undefined genotypes and only three estimated allele frequencies of each marker in the A and genotypes identified as P.  gondouinii. The coordinates F genomes in order to detect A and F genome-specific on the first axis of the factorial analysis allow to clearly markers. Let qAi be the frequency of the allele 0 in the P. assign a posteriori each genotype to one species, that is, to avium genome at the locus i and let qFi be the frequency of the allele 0 in the F genome of P. cerasus brought by P. fruticosa through allopolyploidization. Thus, markers Prunus avium L. for which qFi ¼ 1 are specific to the A genome (no amplification in F genome because the allele 0 is fixed), 0.8 Prunus cerasus L. whereas markers for which qAi ¼ 1 are specific to the F Prunus x gondouinii Pall. 0.6 genome markers. Assuming that P. avium is at Hardy– Unassigned genotypes

Weinberg equilibrium,pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi the allelic frequency qAi was ^ 0.4 estimated by qAi ¼ ½0Šavium; where [0]avium is the frequency of the phenotype [0] in P. avium. Several 0.2 assumptions have to be made to carry on the estimation of qFi , that is, to identify specific markers. We first 0 assumed that both P. avium and P. cerasus are weakly -0.2 spatially structured and that their allelic frequencies 2Dim (6.2%) are still close to their initial state (ie before the allopolyploidization event). Under these hypotheses, -0.4 qAi in P. avium and in P. cerasus should be closely related. Thus, in P. cerasus, ½0Š ¼ ^q2 ^q2 , where -0.6 cerasus Fi Ai [0]cerasus is the frequency of the phenotype [0] in this qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi -0.8 species. For each locus, qFi was estimated as 2 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 ^qF ¼ ½0Š =^q . Many ^qF were higher than 1, i cerasus Ai i Dim 1 (31.5%) indicating that the real frequency qAi in P. cerasus is lower than in P. avium for these loci; these markers are Figure 2 Coordinates of correspondence analysis on 75 AFLP markers of 114 genotypes on the two first axes. Symbols indicate a clearly A genome specific and we considered that ^qFi ¼ 1. For each marker, we plotted the frequencies ^q versus ^q . priori assignation genotypes, based on morphological descriptors. Fi Ai The two first axes represent 36.7% of total variation. Each genotype with coordinates on the first axis lower than À0.3 is a posteriori Genetic diversity of the A genome in P. avium and P. classified in P. avium species; each genotype with coordinates up to cerasus: P. cerasus comes from crosses between P. avium 0.3 is P. cerasus genotype, and each genotype with coordinates and P. fruticosa. We wanted to know whether the genetic between –0.3 and 0.3 is P.  gondouinii genotype.

Heredity Genetic relationships between three cherry species M Tavaud et al 636 classify unknown genotypes in one species and to correct 1 misidentifications as summarized in Table 1. Each 0.9 Class 1 sample having a coordinate on the first axis lower than À0.3 was classified as P. avium species; each sample with 0.8 a coordinate higher than 0.3 was classified as P. cerasus 0.7 Class 2 genotype. Between these two groups, intermediate 0.6

 F genotypes are classified in P. gondouinii species. With 0.5 the 75 markers used, we detected five pairs of identical q genotypes: one pair of P. cerasus, two pairs previously 0.4 classified in P. avium (a posteriori P.  gondouinii) and two 0.3 pairs previously classified in P.  gondouinii (a posteriori P. 0.2

avium and P. cerasus). These five pairs have thus the same 0.1 Class 3 coordinates on the two first axes of the factorial analysis Class 4 and are consequently merged in Figure 2. 0 0 0.2 0.4 0.6 0.8 1 The second axis, corresponding to 6.2% of total inertia, qA gives information on intraspecific variation. An analysis of the three species separately did not show clear genetic Figure 3 Frequencies of the absent allele (0) for the 75 AFLP structure in P. avium or in P. cerasus. However, among the markers in P. avium genome (qA) versus in F genome in P. cerasus (qF). Class 1 contains AFLP markers specific to the A genome, class 2 P. Â gondouinii genotypes, we noticed three distinct contains AFLP markers specific to the F genome. Class 3 contains groups also distinguishable in the global factorial AFLP markers fixed in the F genome of P. cerasus and polymorphic analysis, one composed of Italian genotypes previously in P. avium and class 4 is composed of five markers always found in classified as P. avium and the two others were French the F genome of P. cerasus species but never in P. avium species. varieties (data not shown). Using the a posteriori classification, 61 markers (81.33%) of the 75 used are polymorphic in P. avium, 55 markers (73.33%) are sample, that is, they are always present in the F polymorphic in P. cerasus and 52 (69.33%) are poly- genome and always absent in the A genome. morphic in P. Â gondouinii. Thus, the allelic richness is  Class 5: When 0o^qFo1 and 0o^qAo1, these 21 highest in P. avium. markers show polymorphism on both genomes. With the A- and F-specific markers, the P. Â gondouinii Variation of IBN among the different species genotypes analyzed are separated in three and two When using this a posteriori species classification, the groups, respectively, indicating that there could have species means of IBN are very different between species: been at least three origins of the A genome and two 28.8 and 50.2 for P. avium and P. cerasus, respectively, and origins of the F genome in their hybrid species. This 36.6 for the simulated autotetraploid species (AAAA), far suggests that crosses between P. avium and P. cerasus may less than values from P. cerasus (data not shown). The occur recurrently. difference in IBN between this ‘simulated autotetraploid’ and P. cerasus means that the number of markers observed in P. cerasus may not be explained by doubling Comparison of genetic diversity in A genome between chromosome number alone but may be also due to P. avium and P. cerasus additional alleles, specific to the second genome of P. Gene diversity (He) is 0.319 in P. avium and 0.173 in P. cerasus, that is, the genome coming from P. fruticosa. This cerasus using only the 25 genome A-specific markers. The supports the hypothesis of the allopolyploid origin of P. difference Hei aviumÀHei cerasus is positive for 21 markers cerasus and suggests that indeed specific alleles come among the 25 used, and is therefore highly significant from the F genome. The P. Â gondouinii genotypes have using a sign test (Po0.001). a mean IBN value of 38.6, which is intermediate between P. avium and P. cerasus IBNs. Discussion Identification of genome-specific markers To detect genome-specific markers, we plotted ^q versus Species assignation F Using AFLP markers, the three species P. avium, P. cerasus ^qA, the estimated allele frequencies of the 75 markers in the F and the A genomes, respectively. It was then and P. Â gondouinii are clearly distinguished according to possible to distinguish five classes of markers (Figure 3): their coordinates on the first axis of the correspondence analysis. This separation was confirmed using 725  Class 1: When ^qF  1 and 0o^qAo1, these markers are genotypes of P. avium, representing the whole natural always absent in the F genome and polymorphic in the range of this species (data not shown). Beaver et al (1995) A genome; these 25 markers are called A-specific analyzed seven isozymes markers on 36 P. cerasus, P. markers. avium and P. fruticosa and their hybrids. They identified  Class 2: When ^qA ¼ 1 and 0o^qFo1, these markers are F-specific markers and showed that the first principal always absent in the A genome and are polymorphic coordinate of the PCO clearly separates the diploid from in the F genome. These 14 markers are F-specific. the tetraploid species, but we were not able to distin-  Class 3: When ^qF ¼ 0 and 0o^qA o1, these 10 markers guish each species in the tetraploid species group. are fixed in the F genome and they are polymorphic in AFLP markers are good tools for cherry species the A genome. identification, especially to clearly distinguish tetra-  Class 4: There are five markers for which ^qF ¼ 0 and ploids. Microsatellite markers have been recently deve- ^qA ¼ 1. They are F genome diagnostic markers in our loped in P. avium species (Schueler et al, 2003 ) or

Heredity Genetic relationships between three cherry species M Tavaud et al 637 developed in other Prunus species and used in P. avium using a representative sample of whole P. cerasus genetic (Wunsch and Hormaza, 2002). These microsatellite diversity. markers are codominant and should be useful for the The original habitat of wild P. cerasus is currently study of genetic diversity in specific components of the unknown and is difficult to determine because many genome, particularly if primers which only amplify a feral forms have escaped from orchards. It is not product from one particular genome can be found. possible, therefore, to determine a priori which popula- Several phenomena could explain the number of tions of P. avium contributed to the A genome of P. misidentifications in our sample. Many cherry genotypes cerasus. The national ranges of P. avium and P. fruticosa in this study were old cherry trees, maintained in overlap around the Caspian Sea and Asia Minor; wild P. national orchards and only few morphological traits cerasus could be assumed to have appeared in this were observed for species assignment. Moreover, the sympatric area (De Candolle, 1883; Hedrick, 1915). The distinction between P. avium and P. cerasus species may genome-specific markers detected in our study could be not be easy, especially during winter (Hillig and Iezzoni, very useful for studying the evolution of one genome 1988). The morphological continuum between P. avium among different species with different ploidy levels. We and P. cerasus could be due to the range of cold selection tried to use A genome-specific markers to study the among geographical areas (Iezzoni and Mulinix, 1992) origin of the A genome of P. cerasus, that is, to determine and to complex back-crosses occurring with the local which population(s) of P. avium has contributed to the progenitor species (Beaver et al, 1995). The identification constitution of P. cerasus. Our inability to unambiguously of P. Â gondouinii is still more difficult because of its define the origin of the A genome of P. cerasus is probably phenotypic variability, which ranges from P. cerasus to due to the low number of P. avium genotypes in our P. avium morphologic traits. sample, which is not large enough to represent the whole Specific markers of the F genome have been identified genetic diversity of this species. Moreover, the low using the frequency of the different markers in both geographic differentiation in P. avium makes this deter- genomes of P. cerasus. The diagnostic markers (desig- mination difficult (Mariette et al, 1997; Mohanty et al, nated 4 in Figure 3), always present in F genome of P. 2001). cerasus and always absent in P. avium genome in our The formation of P. cerasus and P. Â gondouinii species sample, will now be helpful to detect if a genotype both involved unreduced P. avium gametes. These contains the F genome. These could be used for species abnormalities during meiosis occur at low frequencies discrimination, especially to sort P. avium from and could be affected by genetic and environmental P. Â gondouinii. It would therefore be beneficial to factors (Iezzoni and Hancock, 1984, 1996; Ramsey and transform the AFLPs into SCAR markers, to allow Schemske, 1998). Nevertheless, we found two different efficient analysis of a large number of genotypes (Xu groups of P. Â goundouinii using A genome-specific et al, 2001). markers. This suggests that the frequency of unreduced Genetic resources characterization, management and gamete production is sufficient to have led to two conservation are essential for future breeding programs. distinct hybridization events. As species assignation based only on morphological traits could lead to misidentifications, we demonstrated that this assignation could efficiently be carried out with Detection of genome-specific markers AFLP markers. Moreover, the same AFLP analysis could There are some limitations in our definition of genome- be used to fingerprint varieties and then to assess the specific markers. Our sets of specific markers are stable genetic diversity present in germplasms, to detect with many P. avium samples (unpublished data), but the redundant genotypes in order to constitute and conserve classification of markers was based on some unproven a core collection. assumptions. The genetic diversity of P. avium has been shown to be weakly spatially structured (Mariette et al, 1997; Mohanty et al, 2001), but we have no data about Origin of P. cerasus impacts on the genetic diversity of any population The difference in distribution of the IBN in P. cerasus bottleneck which could have accompanied the consti- compared to a simulated theoretical autotetraploid tution of P. cerasus, nor the genetic drift since this (AAAA) indicates that P. cerasus is an allotetraploid constitution. Moreover, Beaver et al (1995) suggested containing another genome that is differentiated from the that P. fruticosa, P. cerasus and P. avium share a common A genome. This fact is in agreement with previous gene pool and/or are continually sharing alleles through studies indicating that the P. cerasus origin was a cross introgression. between P. avium and P. fruticosa (Olden and Nybom, In each species, the genotypes studied in our sample 1968). The presence of 46 genome-specific markers and did not represent the whole intraspecific genetic diver- only 29 markers detected in both the A and F genomes sity. For example, P. cerasus analyzed in our sample were demonstrates that these two genomes are quite divergent collected in French orchards and could originate from the but also that some regions of these genomes are same pool of progenitors. That is probably why the conserved, which is in accordance with the segmental correspondence analysis, with only A and F genome- allopolyploidy proposed by Beaver and Iezzoni (1993). specific markers, respectively, did not reveal the clear The reduction of genetic diversity observed in the genetic structure in P. cerasus, which has previously been genome A of P. cerasus in comparison with P. avium is reported (Iezzoni and Hancock, 1996). probably due to the constitution of P. cerasus involving To confirm that F genome-specific markers come from only a part of P. avium genetic diversity. However, in P. fruticosa, samples of this species should be genotyped. this work, mainly French sour cherry trees were It would then be easier to classify each set of markers analyzed. Therefore, these results have to be confirmed according to their presence or absence in each diploid

Heredity Genetic relationships between three cherry species M Tavaud et al 638 species and in tetraploid species carrying the same Iezzoni AF, Schmidt H, Albertini A (1990). Cherries (Prunus). In: genome: a procedure which has been successful in Moore JN, Bellington Jr JR (eds) Genetic Resources of Temperate Gossypium (Abdalla et al, 2001). Fruit and Nut Crops 1. International Society of Horticultural A large number of crosses between species could occur Science: Wageningen, pp 109–173. within the genus Prunus and particularly within the Iezzoni AF, Hancock AM (1984). A comparison of pollen size in subgenus Cerasus (Iezzoni et al, 1990; Webster, 1996). sweet and sour cherry. Hortscience 19: 560–562. Jones CJ, Edwards KJ, Castaglione S, Winfield MO, Sala F, van Backcrosses between P. cerasus and its progenitors exist. de Wiel C et al (1997). Reproducibility testing of RAPD, AFLP The four species P. avium, P. fruticosa, P. cerasus and and SSR markers in plants by a network of European P.  gondounii are often involved in many crosses with laboratories. Mol Breeding 3: 381–390. other species, even those from other botanical sections. Krahl KH, Lansari A, Iezzoni AF (1991). Morphological For example, P. fruticosa could cross with Prunus variation within a sour cherry collection. Euphytica 52: canescens (Rehder, 1947). It would be very interesting to 47–55. study all the Prunus subgenus Cerasus species, in order Krussmann G (1978). Manual of Cultivated Broadleaved Trees and to analyse the genetic relationships between them and to Shrubs, Vol. III. PRU-Z.B.T. Batsford Ltd: London, pp 18–58. have a better knowledge of the crosses involved. Mariette S, Lefranc M, Legrand P, Taneyhill D, Frascaria-Lacoste N, Machon N (1997). Genetic variability in wild cherry populations in France. Effects of colonizing processes. Theor Acknowledgements Appl Genet 94: 904–908. We thank those who supplied plant samples, especially R Mohanty A, Martin JP, Aguinagalde I (2001). A population Saunier, F Santi, M Leroux, M Dechaume, B Heinze, J genetic analysis of chloroplast DNA in wild populations of Prunus avium L. in Europe. Heredity 87: 421–427. Kleinschmidt, members of the Cytofor European pro- Nei M (1978). Estimation of average heterozygosity and genetic gram and members of the European and French net- distance from a small number of individuals. Genetics 89: works for Prunus genetic resources. We also thank J 583–590. Ronfort and T Bataillon for their comments on the Olden EJ, Nybom N (1968). On the origin of Prunus cerasus L. manuscript. This work was supported by a grant from Hereditas 70: 3321–3323. the ‘Association Danone pour les Fruits’, by a CEDRE Ramsey J, Schemske DW (1998). Pathways, mechanisms, and project no. 99 Fb F12/L14. rates of polyploid formation in flowering plants. Ann Rev Ecol Syst 29: 467–501. Rehder A (1947). Manual of Cultivated Trees and Shrubs, 2nd edn. References Macmillan Company: New York, pp 452–481. Santi F, Lemoine M (1990a). Genetic markers for Prunus avium Abdalla AM, Reddy OUK, ElZik KM, Pepper AE (2001). L. 2: clonal identifications and discrimination from P. cerasus Genetic diversity and relationships of diploid and tetraploid and P. cerasus  P. avium. Ann For Sci 47: 219–227. cottons revealed using AFLP. Theor Appl Genet 102: 222–229. Aggarwal RK, Brar DS, Nandi S, Huang N, Khush GS (1999). Santi F, Lemoine M (1990b). Genetic markers for Prunus avium Phylogenetic relationships among Oryza species revealed by L. 1: inheritance and linkage of isozyme loci. Ann For Sci 47: AFLP markers. Theor Appl Genet 98: 1320–1328. 131–139. Badenes ML, Parfitt DE (1995). Phylogenetic relationships of SAS Intitute I (1985). SAS User’s Guide: Statistics. Version 8 cultivated Prunus species from an analysis of chloroplast (Available at http://www.zi.unizh.ch/software/unix/stat- DNA variation. Theor Appl Genet 90: 1035–1041. math/sas/sasdoc/stat/). Baird WV, Ballard RE, Rajapakse S, Abbott AG (1996). Progress Saunier R, Claverie J (2001). Le cerisier: e´volution de la culture in Prunus mapping and application of molecular markers to en France et dans le monde. Point sur les varie´te´s, les germplasm improvement. Hortscience 31: 1099–1106. portegreffe. Le fruit belge 490: 50–62. Beaver JA, Iezzoni AF (1993). Allozyme inheritance in tetra- Schueler S, Tusch A, Schuster M, Ziegenhagen B (2003). ploid sour cherry (Prunus cerasus L.). J Am Soc Hortic Sci 118: Characterization of microsatellites in wild and sweet cherry 873–877. (Prunus avium L.). Markers for individual identification and Beaver JA, Iezzoni AF, Ramm CW (1995). Isozyme diversity in reproductive processes. Genome 46: 95–102. sour, sweet, and ground cherry. Theor Appl Genet 90: 847–852. Schuster M, Schreiber H (2000). Genome investigation in sour Brettin TS, Karle R, Crowe EL, Iezzoni AF (2000). Chloroplast cherry, P. cerasus L. Acta Horticulture 538: 375–379. inheritance and DNA variation in sweet, sour, and ground Viruel MA, Messeguer R, Vicente MCd, Garcia-Mas J, cherry. J Hered 91: 75–79. Pingdomenech P, Vargas F et al (1995). A linkage map De Candolle A (1883). L’origine des plantes cultive´es. Diderot with RFLP and isozyme markers for almond. Theor Appl Multime´dia: Paris, pp 214–221 (edn 1998). Genet 91: 964–971. Faust M, Suranyi D (1997). Origin and dissemination of cherry. Vos P, Hogers R, Bleeker M, Reijans M, van de Lee T, Hornes M Hortic Rev 19: 263–317. et al (1995). AFLP: a new technique for DNA fingerprinting. Hancock AM, Iezzoni AF (1988). Malate dehydrogenase Nucleic Acids Res 23: 4407–4414. isozyme patterns in seven Prunus species. Hortscience 23: Webster AD (1996). The taxonomic classification of sweet and 381–383. sour cherries and a brief history of their cultivation. In: Hedrick UP (1915). The history of cultivated cherries. In: Webster AD, Looney NE (eds) Cherries: Crop Physiology, Hedrick UP, Howe GH, Taylor OM, Tubergen CB, Wellington Production and Uses. Cab International: Wallingford, R (eds) The Cherries of New York. JB Lyon company: Albany, pp 3–23. NY, pp 39–64. Wunsch A, Hormaza JI (2002). Molecular characterization of Hillig KW, Iezzoni AF (1988). Multivariate analysis of a sour sweet cherry (P. avium L.) genotypes using peach (Prunus cherry germplasm collection. J Am Soc Hortic Sci 113: 928–934. persica L. Batsch) SSR sequences. Heredity 89: 56–63. Iezzoni AF, Hancock AM (1996). Chloroplast DNA variation in Xu M, Huaracha E, Korban SS (2001). Development of sour cherry. Acta Hortic 410: 115–120. sequence-characterized amplified regions (SCARs) from Iezzoni AF, Mulinix CA (1992). Variation in bloom time in a amplified fragment length polymorphism (AFLP) markers sour cherry germplasm collection. Hortscience 27: 1113–1114. tightly linked to the Vf gene in apple. Genome 44: 63–70.

Heredity