Euphytica 136: 125–137, 2004. © 2004 Kluwer Academic Publishers. Printed in the Netherlands. 125 Amplified fragment length polymorphism for variety identification and genetic diversity assessment in oleander (Nerium oleander L.) Ezio Portis1 , Cinzia Comino1, Anna Lenzi2 , Piero Lombardi2 , Romano Tesi2 & Sergio Lanteri1,∗ 1 Di.Va.P.R.A. – Plant Genetics and Breeding, University of Turin, via L. da Vinci 44, 10095 Grugliasco (TO), Italy; University of Florence, Piazzale delle Cascine 18, 50144 Florence, Italy; (∗ author for correspondence, e-mail: sergio.lanteri@unito.it) 2 DISAT, Received 25 August 2003; accepted 6 January 2004 Key words: AFLP, Apocynaceae, diversity, DNA fingerprinting, Nerium oleander L. Summary Oleander is a Mediterranean evergreen shrub found along watercourses, gravelly places and damp slopes. It is grown widely as an ornamental for its abundant and long-lasting flowering as well as its moderate hardiness. Genetic relatedness among 71 accessions, including commercial varieties, different sources of the same varieties, and selections from the wild were investigated using amplified fragment length polymorphism (AFLP). Nine primer combinations yielded a total of 603 bands of which 241 were polymorphic. Genetic similarities among accessions were calculated according to Jaccard’s Similarity Index and used to construct a dendrogram based on the unweighted pair group method using arithmetic averages. Our results show that the AFLP technique, which can simultaneously and assay a large number of loci randomly distributed in the genome, is much more informative on the genetic relationship and origin of accessions than the limited number of morphological characters conventionally used for variety discrimination. Up to about 9% molecular genetic differentiation was detected among morphologically indistinguishable provenances of the same variety; this can be partly attributed to scoring error but mainly to somatic variation occurring during vegetative propagation. On the other hand lower genetic distance values were sometimes found among varieties which differ in morphological characters and are thus commercialised with different names. The possibility of considering the amount of genetic variation within a variety as the threshold value for discrimination of initial varieties and essentially derived varieties is discussed. Introduction Oleander (Nerium oleander L., Family Apocynaceae) is a Mediterranean evergreen shrub characteristic of watercourses, gravelly places and damp slopes. It is widely grown as an ornamental in warm temperate and subtropical regions, due to its abundant and longlasting flowering and moderate hardiness (Kingsbury, 1964; Hardin & Arena, 1974). It is used for screens, hedging along highways, planting along beaches and in urban areas as, by removing suckers and leaving just a few stems, it can also be formed into very attractive small trees. In Northern regions it may be grown as an indoor or patio plant. Oleander has flexible branches with green, smooth bark eventually turning to dark grey. Cut or broken branches exude a thick, white sap (Font-Quer, 1979; Schvartsman, 1979; Lampe & McCann, 1985; Pearn, 1987). The leaves are 5 to 20 cm long, narrow, acuminated or acute in the apex, shortly petiolate, with a coriaceus dark-green blade. Some cultivars have white or yellow variegated leaves. Flowers are produced in terminal heads and their colours vary from deep to pale pink, lilac, carmine, purple, salmon, apricot, copper, orange, yellow and white (Huxley, 1992). Each flower is about 5 cm in diameter with five petals, although some cultivars have double flowers. The fruit consists 126 Table 1. Some morphological characters of the 51 varieties and five selections from the wild in study. D: Diameter; W: Width; SPAD (Soil Plant Analysis Development) chlorophyll levels detected in leaves Variety Album Plenum Algiers1 Alsace Altini Angiolo Pucci1 Arad1 Arizona1 Aurora1 Biancaneve1 Bonfire Capraia Commandant Barthelemy Dimona1 Elat1 Elfo1 Emilie Fiesta Pienk1 Foliis Variegata Hardy Red Isle of Capri1 Italia Jannoch Luteum Plenum Madame Leon Blum Magaly Margaritha Maria Gambetta Maurin des Maures1 Minouche (Ville d’Hyeres)2 Mishna1 Mont Blanc Mrs. Roeding Nana Rosso1 Nomade1 Palermo selection A Palermo selection B Palermo selection D2 Palermo selection E Palermo selection F Papà Gambetta1 Petite Pink1 Petite Red (Maravenne)2 Petite Salmon2 Petite White2 Pink Beauty Professeur Granel Corolla Colour Type D (mm) W (mm) Leaf SPAD White Red White with a pink hue Red Ivory yellow Pink Ivory yellow with a pink hue Pink White Fuchsia Pink Pink Fuchsia pink-Red Pink Pink White with a pink hue Pink Pink Pink Fuchsia pink-Red Pale yellow Fuchsia pink-Red Red Pale yellow Pink Pale pink Pink-fuchsia Yellow Fuchsia pink Fuchsia pink Pale pink White Pale salmon pink Pink with dark margins Pink Pink White Red Yellow Pink Pink-Red Pale pink Red Pale salmon pink White Pale pink Fuchsia pink Double Single Single Single Single Single Single Single Single Single Single Double Single Single Single Single Single Double Single Single Single Single Double Single Single Single Single Single Single Single Double Double Single Single Single Double Single Single Double Single Single Single Single Single Single Double 55.8 ± 1.27 51.0 ± 5.90 60.3 ± 1.17 58.9 ± 1.06 62.3 ± 1.67 38.4 ± 2.22 52.4 ± 1.71 57.4 ± 1.90 50.7 ± 4.33 78.1 ± 2.41 51.7 ± 7.64 67.2 ± 2.02 53.8 ± 2.14 48.8 ± 0.84 51.1 ± 1.07 60.3 ± 1.84 60.3 ± 3.2 59.8 ± 1.62 55.4 ± 1.28 45.0 ± 0.4 56.2 ± 1.06 54.6 ± 1.95 55.6 ± 1.68 70.6 ± 1.64 64.6 ± 1.64 56.0 ± 0.69 69.6 ± 0.59 55.2 ± 0.62 46.6 ± 1.29 36.8 ± 0.96 59.4 ± 1.90 61.4 ± 0.95 43.6 ± 0.69 46.2 ± 2.36 56.7 ± 2.4 66.1 ± 2.21 45.1 ± 0.13 49.2 ± 0.41 62.9 ± 1.30 63.6 ± 1.31 44.3 ± 1.26 48.7 ± 0.84 39.0 ± 0.50 41.7 ± 1.5 63.6 ± 1.46 49.7 ± 1.58 24.8 ± 0.78 13.2 ± 1.17 20.4 ± 0.59 18.9 ± 0.80 20.1 ± 0.48 10.0 ± 0.05 14.9 ± 0.51 20.2 ± 1.35 17.8 ± 2.58 23.8 ± 0.11 18.5 ± 3.54 30.3 ± 2.22 13.1 ± 0.96 13.7 ± 2.33 19.2 ± 0.84 23.0 ± 0.51 23.0 ± 0.9. 25.8 ± 2.13 21.4 ± 0.44 15.0 ± 0.1 20.4 ± 0.44 19.9 ± 1.31 27.1 ± 2.56 23.9 ± 0.68 22.3 ± 0.69 22.6 ± 0.11 21.0 ± 0.33 18.3 ± 0.19 15.4 ± 0.30 10.7 ± 0.58 28.8 ± 2.47 27.3 ± 1.53 15.6 ± 1.71 16.2 ± 1.41 18.3 ± 0.33 30.1 ± 2.55 15.7 ± 0.67 16.0 ± 0.54 29.1 ± 0.56 22.0 ± 0.33 16.4 ± 0.73 16.9 ± 1.28 10.3 ± 1.20 12.3 ± 0.6 23.0 ± 0.67 17.8 ± 1.75 73.1 ± 3.38 52.8 ± 1.99 73.0 ± 1.70 58.2 ± 2.25 68.5 ± 1.53 72.2 ± 3.68 68.0 ± 1.85 67.2 ± 3.81 67.3 ± 9.45 72.4 ± 0.63 65.7 ± 2.59 57.7 ± 3.13 62.4 ± 0.29 61.6 ± 6.09 57.8 ± 2.69 64.7 ± 2.08 55.5 ± 4.72 79.3 ± 1.68 60.8 ± 2.37 55.0 ± 1.18 60.5 ± 5.27 61.5 ± 4.95 69.8 ± 6.20 81.5 ± 0.92 61.1 ± 4.58 51.1 ± 9.43 56.0 ± 10.21 60.9 ± 0.44 57.7 ± 1.55 60.4 ± 1.13 68.3 ± 3.76 61.7 ± 1.92 61.6 ± 6.09 54.6 ± 2.75 58.6 ± 2.04 50.9 ± 2.56 58.0 ± 1.19 60.2 ± 0.31 63.9 ± 5.59 67.9 ± 1.78 55.1 ± 0.60 49.8 ± 2.43 49.8 ± 3.25 55.5 ± 0.81 69.8 ± 3.62 67.0 ± 0.85 127 Table 1. Continued Variety Ré D ‘JR 95-3’1 Rosa Bartolini1 Roseum Plenum Rosy Rey1 Sausalito1 Sister Agnes Soleil Levant Souvenir d’August Royer Suor Luisa Tito Poggi Corolla Colour Type D (mm) W (mm) Leaf SPAD Fuchsia pink Pink with dark margins Pink Pale pink Ivory yellow with pink margins White Dark salmon pink Pale pink Red Pink Single Single Double Single Single Single Single Double Single Single 67.6 ± 0.30 56.7 ± 1.44 62.5 ± 2.43 44.2 ± 2.45 47.7 ± 1.41 61.3 ± 1.53 67.4 ± 1.31 75.0 ± 4.07 57.1 ± 1.61 71.8 ± 3.95 21.2 ± 0.29 15.9 ± 0.87 29.5 ± 1.39 11.9 ± 0.48 12.6 ± 0.71 22.6 ± 1.24 20.6 ± 0.73 34.0 ± 2.78 19.7 ± 0.33 24.1 ± 0.22 51.3 ± 0.70 65.8 ± 1.77 54.1 ± 2.15 52.6 ± 4.47 70.8 ± 2.86 56.2 ± 1.21 76.2 ± 1.97 52.6 ± 2.68 68.2 ± 2.91 64.1 ± 2.09 1 compact habit. 2 dwarf. of a narrow follicle 7.5 to 17.5 cm long which opens to disperse fluffy seeds. Oleander can be propagated by seed (Pagen, 1988) but, being allogamous and highly heterozygous, it shows great variability in seedling populations. Growers generally use cuttings. Variety identification is mainly based on flower colour and shape, but other discriminating characters are presence of foliage variegation and growth habit. Naming and identifying oleander varieties is difficult, due mainly to sale of material under unreliable names. Thus an accurate method for their identification and characterization is necessary. Recent developments in DNA marker technology provide means for cultivar fingerprinting as well as for assessing genetic diversity and phylogenetic relationships (Ude et al., 2002). The AFLP technique, which is based on selective amplification of restriction fragments from a digest of total genomic DNA, has several advantages over other marker systems currently in use (Vos et al., 1995; Reeves et al., 1998; Ridout & Donini, 1999). It does not require previous knowledge of the species genome, produces a large number of informative polymorphic markers per primer pair, is highly sensitive, requires small amounts of DNA and has proved to be robust, reliable and reproducible (Mueller & Wolfenbarger, 1999; Hodkinson et al., 2000, 2002). To our knowledge, DNA markers have not until now been used to analyse the oleander genome. The objectives of the present study were to evaluate the usefulness of AFLP in differentiating oleander varieties, and to determine genetic relationships in a sample of 71 accessions representative of the most common Table 2. List of the eight varieties of oleander including different provenances Variety Accession Provenances Papà Gambetta I II III IV V VI I II III IV I II III I II I II I II I II I II France Latium (Italy) Marches (Italy) Liguria∗ (Italy) Liguria∗ (Italy) Sicily (Italy) Latium (Italy) France Marches (Italy) Liguria (Italy) Latium (Italy) Marches (Italy) Liguria (Italy) Marches (Italy) Tuscany (Italy) France Marches (Italy) Latium (Italy) Tuscany (Italy) Marches (Italy) France Latium (Italy) Tuscany (Italy) Maria Gambetta Luteum Plenum Emilie Magaly Pink Beauty Tito Poggi Madame Leon Blum ∗ Accessions IV and V of Papà Gambetta came from the same region, but from different nurseries. commercial varieties, provenances within the same variety, and selections from the wild. 128 Materials and methods Plant material The accessions under study are maintained at DISAT, University of Florence (Lenzi et al., 1999; Lenzi & Tesi, 2000). The collection contains 51 varieties commercialised by Italian and French nurseries as well as 5 Sicilian selections obtained from the wild (Table 1). Eight varieties included different provenances as reported in Table 2. Seventy one accessions were included in our analysis. For each accession the growth habit (i.e. vigorous, compact or dwarf) and the following morphological characters were recorded: corolla colour (measured using a portable colorimeter NR-3000, Nippon Denshoku), type (double or single), diameter and width; chlorophyll levels were measured in three leaves of at least two plants using the portable equipment SPAD502 (Minolta). Examples of flower form and colour in oleander varieties are shown in Figure 1. DNA extraction and AFLP analysis One young leaf was collected from three plants per accessions and the three pooled leaves used for DNA extraction according to Lanteri et al. (2001). The AFLP protocol was essentially that of Vos et al. (1995) with minor modifications (Lanteri et al., 2003). Restriction and ligation were done concurrently by adding 5 µl extracted DNA (400–500 ng DNA) to 45 µl buffer (10 mM Tris-HCl pH 7.5; 10 mM MgAc, 50 mM KAc) containing 5 units EcoRI, 5 units MseI, 2 units T4 DNA ligase (New England BioLabs, Beverly, MA), 5 pmol EcoRI adapter, 50 pmol MseI adapter and 0.2 mM ATP. The mixture was then incubated at 37◦ for 4h and diluted 10 times in 0.1× TE (1 mM Tris-HCl, 0.1 mM EDTA pH 8). We used two consecutive PCRs to selectively amplify the EcoRI-MseI DNA fragments. The preselective amplification (first PCR) was performed using 5 µl of the above mentioned diluted mixture added to a 15 µl mixture giving a final concentration of 10 mM Tris-HCl pH 8.3, 50 mM KCl, 1.5 mM MgCl2 , 0.2 mM of each dNTPs (Sigma-Aldrich), 40 ng of EcoRI and MseI adapter-directed primers, each possessing a single selective base (EcoRI+1 and MseI+1 primers), and 1 unit of Taq polymerase (Promega). PCR reactions were performed with the following profile: 94 ◦ C for 60 s, 25 cycles of 30 s denaturing at 94 ◦ C, 30 s annealing at 55 ◦ C and 60 s extension at 72 ◦ C, ending with 10 min at 72 ◦ C to complete extension. After checking for the presence of a smear of fragments (100–1000 bp in length) by agarose electrophoresis, the amplification product was diluted 40 times in 0.1× TE. Subsequently selective amplification (second PCR) was carried out using primers with three selective nucleotides. Initially, 16 primer pairs, originated by the combination of 4 EcoRI primers and 4 MseI primers, were tested in four oleander accessions. From this pilot study, the following nine primer combinations were selected, on the basis of clearness and reproducibility of electrophoretic patterns, and applied to all samples: E32/M49 (AAC/CAG), E32/M50 (AAC/CAT), E32/M62 (AAC/CTT), E33/M49 (AAG/ CAG), E33/M50 (AAG/CAT), E33/M60 (AAG/CTC), E35/M50 (ACA/CAT), E35/M60 (ACA/CTC), E35/ M62 (ACA/CTT). The EcoRI and MseI adapters and primers were synthesized by Invitrogen Life Technologies. Selective PCR reactions were performed with the following profile: 94 ◦ C for 60 s, 36 cycles of 30 s denaturing at 94 ◦ C, 30 s annealing and 60 s extension at 72 ◦ C, ending with 10 min at 72 ◦ C to complete extension. Annealing was initiated at 65 ◦ C which was then reduced by 0.7 ◦ C for the next 12 cycles and maintained at 56 ◦ C for the subsequent 23 cycles. Reproducibility for each primer pair was checked by running the AFLP protocol at different DNA concentrations; a threshold of 20 ng DNA/µl before digestion was the lowest concentration which avoided appearance of artifacts or disappearance of some bands. Electrophoresis of the PCR product Amplification products were mixed with 15 µl of formamide-dye (98% formamide, 10 mM EDTA, 0.01% w/v bromophenol blue and 0.01% w/v xylene cyanol), denatured at 95 ◦ C for 4 min and separated by electrophoresis on 5% denaturing polyacrylamide sequencing gels (5% acrylamide-7 M Urea 19:1) in 1×TBE buffer. The gels were pre-run for about 30 min before 4.5 µl of the mix was loaded. Gels were run at 110 W for about 2.5 h. Staining Gels were silver stained (Bassam et al., 1991). The gel was fixed in 10% acetic acid for 30 min, washed twice with a large quantity of ultrapure water for 5 min, transferred to a silver impregnation solution (1g L−1 AgNO3 , 0.056% formaldehyde) for 30 min and then rinsed with ultrapure water for 5 s. All steps were 129 Figure 1. Example of flower types and colours among oleander varieties (A). Flower and leaf of ‘Minouche’ (B) and ‘Commandant Barthelemy’ (C). 130 performed with slow agitation on a shaker. Image development was carried out with manual agitation for 1 to 3 min in developer (30 g L−1 Na2 CO3 , 0.056% formaldehyde, 400 µg L−1 sodium thiosulphate). To stop development and fix the gel, 10% acetic acid was added directly to the developing solution, and shaking continued for 2–3 min. The gel was then rinsed briefly in ultrapure water and dried at room temperature. Data scoring and analysis AFLP amplifications were repeated at least twice in order to test their consistency. Electrophoretic patterns were documented using the Gel Documentation System (Quantity One Programme, BioRad). Each PCR product was assumed to represent a single locus and only reproducible polymorphic bands were scored as present (1) or absent (0). AFLP data were evaluated by means of Shannon’s Index, Marker Index and Polymorphic Information Content. Shannon’s Index (H’j ) (Shannon and Weaver, 1949) for each locus was calculated as follows: H’j = –
pi log2pi , where pi is the frequency of the presence or absence of fragment. In order to compare the level of diversity detected by different primer combinations, we partitioned diversity for each primer combinations by subtotalling H’j . Marker Index (MI) was calculated according to Powell et al. (1996) as the product of two functions: Expected Heterozygosity (Hn ) and Effective Multiplex Ratio (EMR). Hn of a locus is defined as: 1 –
pi 2 , where pi is the frequency of the presence or absence of the fragment (band). EMR of a primer was defined as: βn, were β is the percentage of polymorphic loci and n is the number of loci detected per primer (Milbourne et al., 1997). The polymorphic information content (PIC) was calculated by applying the simplified formula for the Expected Heterozygosity (Anderson et al., 1993): PIC = 2f (1-f), were f is the percentage of plants where the marker is present. A binary matrix for cluster analysis was prepared using the NTSYS-pc (numerical taxonomy and multivariate analysis system) version 1.80 package (Rohlf, 1993). Genetic similarity among accessions was calculated according to Jaccard’s Similarity Index (JSI) (Jaccard, 1908) in all possible pair-wise comparisons, using the SIMQUAL (similarity of qualitative data) routine. The JSI was defined as: JSIxy = a/(a+b+c), where a = number of bands shared from individuals x and y, b = number of bands present in x and absent in y, c = number of bands present in y and absent in x; thus, JSIxy = 1 indicates identity between x and y, whereas JSIxy = 0 indicates complete divergence. The JSIs were used to construct a dendrogram using UPGMA (unweighted pair-group method, arithmetic average) through the SAHN (Sequential, agglomerative, hierarchical, and nested cluster analysis) routine. A co-phenetic matrix was produced using the hierarchical cluster system, by means of the COPH routine, and correlated with the original distance matrices for AFLP data, in order to test for agreement between the cluster in the dendrogram and the JSI matrix. Clustering ability test The clustering abilities of nine selected AFLP primer combinations were tested to determine the optimal number of primer pairs needed to discriminate the maximum number of oleander accessions. The primer combination with the highest ability to cluster the 71 accessions was first analysed alone (PC1.J = primer combination 1, based on Jaccard’s Similarity Index) and then in combinations with the other PCs with progressively lower discrimination power. The last combination (PC9.J) comprised all the nine PCs used in the study. Results Morphological characterization Data on morphological characterization of the 51 varieties and five selection from the wild in study are reported in Table 1. Primer selection and AFLP analysis Preliminary tests were conducted using the preamplification products in order to define the conditions that would yield distinct amplified fragments on the sequencing gel. We tested different combinations of selective primers, with three selective nucleotides at the EcoRI end (EcoRI+3) and from one to three selective nucleotides at the MseI end (from MseI+1 to MseI+3). On the whole from 4 to 6 selective nucleotides were tested. Only the primer combinations characterised by 6 selective nucleotides produced scoreable bands and were used in this study; the others resulted in a smear or yielded too many fragments for accurate scoring. 131 Nine of the 16 primer combinations gave clear and reproducible amplification patterns. Among the 7 discarded primer combinations, three amplified too many bands for accurate scoring; although it could be argued that these primer combinations are highly informative, the difference in electrophoretic mobility between bands was very small and increased the risk of misalignment. The other four primer combinations discarded (all of them characterized by the presence of the Eco+ACG primer) were not very informative as they yielded small numbers of bands often not uniformly distributed in the gel. A total of 241 polymorphic bands (39.9% of the total amplified bands), ranging from 40 to 750 bp, were scored (Table 3). The average number of polymorphic bands per primer combination was 26.8 ranging from 22 to 38 per priming pair (Table 3). The Shannon Index, Polymorphic Information Content and Marker Index for each primer combination are also reported in Table 3. Primer combination E35/M60 showed the highest values for H’j and PIC, while primer combination E33/M50 showed the highest value for MI and allowed to distinguish 52 of the 56 varieties and selections from the wild analysed, and 59 of the 71 accessions studied. Primer combination E33/M49, gave the lowest values for H’j , PIC and MI. Five of the nine primer combinations used made possible the detection of 15 unique/distinctive bands (i.e. fragments present in only one accession, Table 3). Eight of the 56 varieties and selections were characterized by distinctive bands (Table 4). An example of an AFLP profile is shown in Figure 2. morphological characters or growth habits usually adopted for varietal identification. Although branch A includes only varieties with compact habit and the dwarf ‘Petit Salmon’, other varieties with compact or dwarf habit were distributed in the other clusters. Moreover although yellow cultivars were mainly included in cluster B1, the yellow flowered ‘Sausalito’ and ‘Luteum Plenum’ were in cluster A and in the B3 out-groups respectively. Interestingly most of the double-flowered varieties were in cluster B2, and three of them in B3 out-groups. The co-phenetic correlation coefficient (r-value) between the data matrix and the co-phenetic matrix for AFLP data was 0.88, suggesting a good fit between the dendrogram and the similarity matrix from which it was derived. Genetic relatedness Many oleander varieties are now available and commercialised, therefore their accurate identification is becoming important; however, at present, classification is based on a limited number of characters, mainly shape, size or colour of the corolla, presence of foliage variegation and growth habit. This appears to be the first report of the use of a DNA-based polymorphism assay to identify genetic differences among oleander varieties, which for commercial purposes are vegetatively propagated. For clonally propagated ornamentals, varietal uniformity and stability are only influenced by somaclonal variation, therefore testing authorities are studying the possibility to apply molecular markers for assessing distinctness, uniformity and stability (DUS) criteria for new varieties, and for the management of reference collections (De Riek, 2001). Furthermore, The JSI values ranged from 0.201 for ‘Rosy Rey’ and ‘Commandant Barthelemy’ to 1.00 for the provenances II and III of variety ‘Papà Gambetta’ (see Table 2). The values among all accessions are available on request from the authors. The dendrogram based on the similarity values generated using UPGMA (Figure 3), shows that accessions ‘Rosy Rey’ and ‘Palermo selection A’ were the most divergent, with respectively an average genetic similarity of about 45 and 49% to the others. The dendrogram separated the other accessions into 4 main branches (A, B, C, D) with branch B being subdivided into two major clusters: B1, B2 and a few out-groups (B3). However, it was not possible to consistently correlate the clustering based on AFLP data with Test for clustering ability of primer combinations The number of variety and accession subsets resulting from the successive clustering analyses with the nine primer combinations are shown in Table 5. The number of oleander varieties clustered by using different primer pair was found to increase from 52 in PC1.J, to 56 in PC4.J. The clustering power reached a maximum point of 70 subsets (accessions) after the AFLP data of the sixth primer combination were added. The addition of the other primer combinations (PC7.J, PC8.J, PC9.J) resulted in minor modification of the dendrogram (data not shown). Discussion 132 Table 3. Summary of AFLP primer combination characteristics. Total number of bands (TNB), number of polymorphic bands (NPB), percentage of polymorphic bands (P%), Shannon index (H’j ), Polymorphic Information Content (PIC), Marker Index (MI), number of different varieties and selections from the wild identified (NV), number of different accessions identified (NA), and number of exclusive bands (NEB) obtained per primer combination Primer combination TNB NPB P% H’j PIC MI NV NA NEB 58 67 75 64 80 66 63 63 67 24 25 27 27 38 24 22 24 30 41.4 37.3 36.0 42.2 47.5 36.4 34.9 38.1 44.8 0.678 0.636 0.651 0.525 0.632 0.603 0.685 0.785 0.663 0.306 0.284 0.289 0.220 0.284 0.266 0.314 0.368 0.297 3.038 2.653 2.805 2.505 5.124 2.326 2.412 3.369 3.990 49 46 49 47 52 50 40 51 51 54 50 52 52 59 55 46 58 55 0 3 0 4 5 1 2 0 0 603 67.0 241 26.8 39.8 0.651 0.292 3.136 48.3 53.4 E32/M49 E32/M50 E32/M62 E33/M49 E33/M50 E33/M60 E35/M50 E35/M60 E35/M62 Total average 15 1.66 Table 4. List of varieties and selections in which exclusive bands were detected; number of exclusive bands (NEB); primer combinations (PCs) Variety NEB PCs Palermo selection A Rosy Rey Palermo selection B Dimona Minouche (Ville d’Hyeres) Papa Gambetta Roseum Plenum Souvenir d’August Royer Total 4 4 2 1 1 1 1 1 15 E33/M49 (2) – E33/M50 – E35/M50 E32/M50 – E33/M50 (2) – E35/M50 E32/M50 – E33/M60 E32/M50 E33/M50 E33/M50 E33/M49 E33/M49 Table 5. Primer combinations, number of polymorphic bands (NPB), number of different varieties and selections from the wild (NV) and number of different accessions (NA) they are able to distinguish Dendrogram Primer combinations NPB NV NA PC1.J PC2.J PC3.J PC4.J PC5.J PC6.J PC7.J PC8.J PC9.J E33/M50 E33/M50. E35/M60 E33/M50. E35/M60. E35/M62 E33/M50. E35/M60. E35/M62. E33/M60 E33/M50. E35/M60. E35/M62. E33/M60. E32/M49. E33/M50. E35/M60. E35/M62. E33/M60. E32/M49. E32/M62. E33/M50. E35/M60. E35/M62. E33/M60. E32/M49. E32/M62. E33/M49 E33/M50. E35/M60. E35/M62. E33/M60. E32/M49. E32/M62. E33/M49. E32/M50 E33/M50. E35/M60. E35/M62. E33/M60. E32/M49. E32/M62. E33/M49. E32/M50. E35/M50 38 62 92 116 140 167 194 219 241 52 53 55 56 56 56 56 56 56 59 61 64 67 69 70 70 70 70 133 Figure 2. AFLP pattern obtained by primer combination E33/M50 in 35 of the 51 oleander varieties analyzed. Arrows show presence of exclusive bands in varieties ‘Papà Gambetta’ and ‘Rosy Rey’. 134 Figure 3. Dendrogram obtained from UPGMA cluster analysis of AFLP data generated by the 9 primer combinations tested. 135 molecular markers might find application in detecting infringements of plant breeders’ rights and help in the discrimination of Essentially Derived Varieties (EDVs). As noted by De Riek (2001), a test based on the molecular genetic relatedness between an initial variety (IV) and an EDV is very informative, as even if both varieties have completely different flower shape or colour, they will share most of their genome. In this study we applied the AFLP technique, since it has proved to be powerful in detecting similarities in the genome of related cultivars and has been applied to the assessment of genetic conformity and for testing essential derivation in numerous ornamental plants (Barcaccia et al., 1999; Loh et al., 1999, Leus et al., 2000; Van Huylenbroeck et al., 2001; Tomkins et al., 2001; Carr et al., 2003). In order to apply the technique, we tested a wide range of primer combinations. The need for this has been previously noted (Qi and Lindhout, 1997; Castiglioni et al., 1999; Lima et al., 2002; Lanteri et al., 2004). Among the primer pairs tested, seven were discarded, due to the not clearly interpretable and not reproducible electrophoretic patterns obtained, while 9 were chosen and applied for molecular characterization. Pejic et al. (1998) reported that 150 polymorphic bands make possible a reliable estimate of genetic similarities among genotypes within the same species; indeed we found that the maximum resolution of our 70 subsets (accessions) clustered by UPGMA analysis was obtained using 6 primer combinations, which made possible the amplification of 167 polymorphic bands. Nevertheless our study was based on the detection of 241 polymorphic bands, which permitted more accurate estimates of the genetic relationships being studied. To obtain unambiguous attribution of accessions to a variety, we characterized each accession by the growth habit and morphological characters usually adopted for varietal identification, and we confirmed that different provenances within the same variety were always indistinguishable. Two accessions, ‘Rosy Rey’, with compact habit and single pink flowers, and ‘Palermo selection A’, with the same flower characters but a more vigorous habit, were highly genetically differentiated from all the others. Indeed, in both of them, we detected four exclusive bands, which might be converted into STS (sequence tagged site) markers of great values for varietal fingerprinting. Interestingly, among Sicilian selections, ‘Palermo selection A’ was the only one which did not cluster with other com- mercial varieties, and this might confirm its derivation from autochthonous instead of naturalised germplasm. The other 69 accessions could be grouped in four main branches of which branch B was further subdivided into two major clusters. On the whole, it was not possible to correlate morphological characters usually adopted for variety identification with the clustering obtained with molecular data. Varieties with different corolla colour or size as well as growth habit were quite uniformly distributed among the clusters. Interestingly varieties with double corolla were always included in cluster B2 and in B3 out-groups; this supports the hypothesis of their different origin and introduction at the end of the 17th century from India (Pagen, 1988), although both single and double corolla types, together with their hybrids, are now present in nature. The weak correlation between morphological and molecular data is not surprising, considering that the limited number of characters used for variety discrimination is encoded by a limited number of genes, which can originate new phenotypes as a consequence of simple mutation events or non-heritable changes: i.e. transposons or epigenetic effects. Vice versa, by means of AFLP markers, we were able to simultaneously and randomly assay a large number of loci in the genome. Provenances within the same variety always clustered together and limited genetic differentiation among them was detected. The range of genetic differentiation was about 3% among the three accessions of ‘Luteum Plenum’ and two accessions of ‘Magaly’ and ‘Tito Poggi’, and even lower (about 2%) for five of the six ‘Papà Gambetta’ and three of the four ‘Maria Gambetta’. However, ‘Maria Gambetta’ accession III was genetically differentiated at about 5% from other provenances of the same variety and an analogous value was detected between the two accessions of ‘Emilie’. Interestingly, ‘Papà Gambetta’ accession V was more genetically similar to ‘Rosa Bartolini’ than to the other accessions within the same variety, from which an average genetic distance of 9% was detected; an analogous value was found between the two accessions of ‘Madame Leon Blum’ and of ‘Pink Beauty’. By comparing AFLP profiles of identical clones and replicate samples we estimated that the scoring error in our analyses was about 2%, which is consistent with that estimated in other studies (Mueller & Wolfenbarger, 1999; Hodkinson et al., 2002); higher values can thus be attributed to somatic variation occurring during vegetative propagation. 136 The lowest genetic similarity among morphologically indistinguishable provenances of the same varieties, i.e. JSI = 0.879 between ‘Papà Gambetta’ accessions V and I, may be considered the threshold value due to somaclonal variation occurring over time. Thus the distance of about 9% between ‘Tito Poggi’, ‘Madame Leon Blum’ and ‘Aurora’, which are phenotypically very similar, suggests that they share the same genetic background and presumably the same origin. Indeed, Pagen (1988) reports that ‘Tito Poggi’ is a selection with darker flowers of ‘Madame Leon Blum’, while for Filippi (1997) states that the two varieties might be retraced to the same variety and are both very similar to ‘Soleil Levant’, which from our data was indeed 9% distant from the others. A distance of about 4% was detected among ‘Roseum Plenum’, ‘Palermo selection F’ and ‘Foliis Variegata’, all of them with double pink flowers but the last differing in the presence of leaf variegation, which can thus be attributed to mutation of a common ancestor; furthermore, distances lower than 9% were found between ‘Magaly’ and ‘Pink Beauty’, both with simple pale pink flowers, as well as between ‘Jannoch’ and ‘Suor Luisa’, both with single red flowers. Notwithstanding its wide popularity and commercialization, there is no Official Variety Register for oleander. Our data demonstrate that when two varieties, although morphologically distinguishable and commercialised with different names, show a molecular genetic differentiation lower or analogous to that detectable among provenances of the same variety, they might be considered as EDVs; at least, molecular markers should function as a strong indication to competing breeders to prove the origin of their new selections. References Anderson, J.A., G.A. Churchill, J.E. Autrique, M.E. Sorells & S.D. Tanksley, 1993. Optimizing parental selection for geneticlinkage maps. Genome 36(1): 181–186. Barcaccia, G., E. Alberini & M. Falcinelli, 1999. AFLP fingerprinting in Pelargonium peltatum: its development and potential in cultivar identification. Journal of Horticultural Science & Biotechnology 74(2): 243–250. Bassam, B.J., G. Caetano-Anolles & P.M. Gresshoff, 1991. Fast and sensitive silver staining of DNA in polyacrylamide gels. Anal Biochem 19: 680–683. Carr, J., M. Xu, J.W. Dudley & S.S. Korban, 2003. AFLP analysis of genetic variability in New Guinea impatiens. Theor Appl Genet 106(8): 1509–1516. Castiglioni, P., P. Ajmone-Marsan, R. Van Wijk & M. Motto, 1999. AFLP markers in a molecular linkage map of maize: codominant scoring and linkage group distribution. Theor Appl Genet 99: 425–431. De Riek, J., 2001. Are molecular markers strengthening plant variety registration and protection? Acta Horticulturae 552: 215–223. Filippi, O., 1997. Guide de reconnaissance des variétés de lauriersroses. Pépinière Filippi, Meze (France). Font-Quer, P., 1979. Plantas Medicinales. Barcelona, Labor SA. Hardin, J.W. & Arena J.M., 1974. Human Poisoning from Native and Cultivated Plants, 2nd edn. Kingsport, Tennessee, Duke, University Press. Hodkinson, T.R., S.A. Renvoize, G. Ní Chonghaile, C. Stapleton & M.W. Chase, 2000. A comparison of ITS nuclear rDNA sequence data and AFLP markers for phylogenetic studies in Phyllostachys (Bambusoideae, Poaceae). J Plant Res 113: 259–269. Hodkinson, T.R., M.W. Chase & S.A. Renvoize, 2002. Characterization of a genetic resource collection for Miscanthus (Saccharinae, Andropogoneae, Poaceae) using AFLP and ISSR PCR. Ann Botany 89: 627–636. Huxley, A., ed-in-chief. 1992. The New Horticultural Society Dictionary of Gardening, vol. 3. London, MacMillan. Jaccard, P., 1908. Nouvelles recherches sur la distribution florale. Bull Soc Vaudoise sci nat 44: 223–270. Kingsbury, J.M., 1964. Poisonous Plants of the United States and Canada. Englewood Cliffs, NJ Prentice Hall. Lampe, K.F. & M.A. McCann, 1985. AMA Handbook of Poisonous and Injurious Plants, American Medical Association, Chicago, Illinois. Lanteri, S., I. Di Leo, L. Ledda, M.G. Mameli & E. Portis, 2001. RAPD variation within and among populations of globe artichoke cultivar ‘Spinoso sardo’. Plant Breed 120: 243–246. Lanteri, S., A. Acquadro, L. Quagliotti & E. Portis, 2003. RAPD and AFLP assessment of genetic variation in a landrace of pepper (Capsicum annuum L.) , grown in North-West Italy. Genet Res Crop Evol 50(7): 723–735. Lanteri, S., E. Saba, M. Cadinu, G.M. Mallica, L. Baghino & E. Portis, 2004. Amplified fragment length polymorphism for genetic diversity assessment in globe artichoke. Theor Appl Genet, in press. Lenzi, A., A. Palandri, A. Bovelli & R. Tesi, 1999. L’oleandro (Nerium oleander L.) per la coltura in contenitore. Colture Protette 9: 101–112. Lenzi, A. & R. Tesi, 2000. La collezione di oleandri del D.I.S.A.T. – Proceedings of the Congress ‘Il germoplasma toscano: tutela e valorizzazione’, pp. 129–136. Leus, L., E. Demuynck & J. De Riek, 2000. Genetic diversity of a collection of rose species and cultivars evaluated by fluorescent AFLP. Med Fac Landbouww Univ Gent 65/3b: 455–459. Lima, M.L.A., A.A.F. Garcia, K.M. Oliveira, S. Matsuoka, H. Arizono, C.L. deSouza Jr. & A.P. de Souza, 2002. Analysis of genetic similarity detected by AFLP and coefficient of parentage among genotypes of sugar cane (Saccharum spp.). Theor Appl Genet 104: 30–38. Loh, J.P., R. Kiew, A. Kee, L.H. Gan & Y.Y. Gan, 1999. Amplified fragment length polymorphism (AFLP) provides molecular markers for the identification of Caladium bicolor cultivars. Ann Botany 84(2): 155–161. Milbourne, D., R. Meyer, J.E. Bradshaw, E. Baird, N. Bonar, J. Provan, W. Powell & R. Waugh, 1997. Comparison of PCRbased marker systems for the analysis of genetic relationships in cultivated potato. Mol Breed 3: 127–136. Mueller, U.G. & L. Wolfenbarger, 1999. AFLP genotyping and fingerprinting. Trends Ecol Evol 14: 389–394. 137 Pagen, F.J.J., 1988. Oleanders. Nerium L. and the Oleander Cultivars – Agricultural University Wageningen, The Netherlandes, pp. 113. Pearn, J., 1987. Oleander poisoning. In: J. Covacevich, P. Davie & J. Pearn (Eds.), Toxic plants & Animals, A guide For Australia, pp. 37–49. Brisbane, Queensland Museum. Pejic, I., P. Ajmone-Marsan, M. Morgante, V. Kozumplick, P. Castiglioni, G. Taramino & M. Motto, 1998. Comparative analysis of genetic similarity among maize inbred lines detected by RFLPs, RAPDs, SSRs and AFLPs. Theor Appl Genet 97: 1248–1255. Powell, W., M. Morgante, C. Andre, M. Hanafey, J. Vogel, S. Tingey & A. Rafalsky, 1996. The comparison of RFLP, RAPD, AFLP and SSR (microsatellite) markers for germplasm analysis. Mol Breed 2: 225–238. Qi, X. & P. Lindhout, 1997. Development of AFLP markers in barley. Mol Gen Genet 254: 330–336. Reeves, G., D. Francis, M.S. Davies, H.J. Rogers & T.R. Hodkinson, 1998. Genome size is negatively correlated with altitude in natural populations of Dactylis glomerata. Ann Botany 82 (suppl A): 99–105. Ridout, C.J. & P. Donini, 1999. Use of AFLP in cereals research. Trends Plant Sci 4: 76–79. Rohlf, F.J., 1993. NTSYS-PC Numerical Taxonomy and Multivariate Analysis System. Version 1.80. Exeter Publ., Setauket, New York. Schvartsman, S., 1979. Plantas Venenosas. Sarvier Sao Paulo. Shannon, C.E. & W. Weaver, 1949. The mathematical theory of communication. In: I.L. Urbana (Ed.), University of Illinois Press. Tomkins, J.P., T.C. Wood, L.S. Barnes, A. Westman & R.A. Wing, 2001. Evaluation of genetic variation in the daylily (Hemerocallis spp.) using AFLP markers. Theor Appl Genet 102(4): 489–496. Ude, G., M. Pillay, D. Nwakanma & A. Tenkouano, 2002. Analysis of genetic diversity and sectional relationships in Musa using AFLP markers. Theor Appl Genet 104: 1239–1245. Van Huylenbroeck, J., E. Coart, F. Janneteau & J. De Riek, 2001. Identification of woody ornamentals by AFLP. Eucarpia Meeting – Ornamental Symposium, ‘Strategies for New Ornamentals’, Melle, Belgium. Vos, P., R. Hogers, M. Bleeker, M. Reijand, T. van de Lee, M. Hornes, A. Fritjers, J. Pot, J. Paleman, M. Kuiper & M. Zabeau, 1995. AFLP: A new technique for DNA fingerprinting. Nucl Acids Res 23: 4407–4414.