Materials and Methods

Plant material.

One hundred four (104) accessions were tested, which represented six plum species (P. salicina, P. simonii, P. ussuriensis, P. domestica, P. cerasifera, and P. spinosa), two related species (P. armeniaca and P. tomentosa), eight plumcots, and six accessions of wild European plum (Table 1); and all were collected from the Chinese National Germplasm Repository for Plums and Apricots located in Xiongyue, Liaoning Province, China, except for six accessions of wild European plum, which were sampled from the Tianshan Germplasm Repository for Wild Fruit Resources located in Xinyuan, Xinjiang Province, China.

Table 1. Accessions of Prunus assayed with ISSR markers in present study.

View Fullscreen

DNA extraction.

Samples were taken from three plants of each accession. DNA extraction was performed using the CTAB method (Doyle and Doyle, 1990) with modification. Briefly, ≈0.1 g of frozen leaves was put into a 1.5-mL Eppendorf tube and ground to a fine powder in liquid nitrogen with a specially designed grinder. The powder was added to 650 μL of extraction buffer (2% CTAB, 1.4 m NaCl, 100 mm Tris-HCl, 20 mm EDTA, 2% BME, 1% PVP) and incubated at 65 °C for 30–60 min. After this mixture had been placed in an ice-bath for 5 min, DNA was extracted once with one volume of chloroform/isoamyl alcohol (24:1, v/v). The DNA suspension was mixed gently and centrifuged at 15,300g _n (4 °C) for 15 min. The top aqueous phase was collected into a new tube, and 2 μL of RNase (10 mg·mL⁻¹) was added, mixed, and then incubated at 37 °C for 30–60 min. DNA was extracted once more with one volume of the chloroform/isoamyl mixture, the top aqueous phase was collected again, and the DNA was precipitated with two volumes of cold ethanol. The DNA pellet was washed three times with 70% ethanol and dissolved in 100 μL of sterile water after the ethanol evaporated completely. DNA quality was checked on a 1% agarose gel, and the concentration was determined by spectrophotometry at 260 nm. A portion of the DNA was diluted to 20 ng·μL⁻¹ for use, and the stock and diluted portions were stored at −20 and 4 °C, respectively.

PCR procedure.

PCR reaction mixtures had a total volume of 20 μL containing 60 ng of DNA template, 0.25 μM primer, 0.2 mm dNTPs, 0.25 unit of Taq DNA polymerase, 3.0 mm MgCI₂, 2 μL 10× buffer [20 mm MgSO₄, 100 mm KCl, 80 mm (NH₄)₂SO₄, 100 mm Tris-HCl, 0.5% NP-40, pH 9.0]. Reaction mixtures were overlaid with 50 μL of mineral oil.

Amplifications were performed on a DNA Thermal Cycler 9700 (PerkinElmer, Wellesley, MA) with the following temperature profile: a pre-denaturation step of 5 min at 94 °C followed by 40 cycles of denaturation, 45 s at 94 °C; annealing, 1 min at 52 °C; extension, 90 s at 72 °C. A final extension was allowed for 5 min at 72 °C.

Amplified products were separated by electrophoresis on 2% agarose gel in 1× TAE buffer under a voltage of 40 V·cm⁻¹ for ≈1.5 h and visualized with staining by ethidium bromide, and then ultraviolet-transilluminated gels were photographed.

Data analysis.

Amplification fragments were scored as present (1) or absent (0). Faintly stained bands were not included in the analysis. Each reaction was repeated twice, hence only reproducible products were included in the study. Based on the number of shared amplification products, data for all 12 primers were assembled into a data matrix. The UPGMA cluster analysis and PCoA were performed using the NTSYS-pc version 2.1 software (Rohlf, 2000) based on the Jaccard coefficient of similarity, and two dendrograms, one with accessions grouped by species and one including all 104 accessions, and a scatter plot were obtained. Mantel tests (Mantel, 1967) were carried out to calculate goodness-of-fit between the cluster analyses and original matrices.

Results

In a pilot experiment, the effects of various levels of magnesium concentrations (1.0, 1.5, 2.0, 2.5, and 3.0 mm), template DNA concentrations (10, 20, 40, 60, and 80 ng), dNTP concentrations (0.1, 0.15, 0.2, 0.25, and 0.3 mm), Taq polymerase contents (0.25, 0.5, 1.0, 1.5, and 2.0 U), primer concentrations (0.1, 0.15, 2.0, 2.5, and 3.0 mm), and annealing temperatures (49, 49.3, 50, 50.8, 52.1, 53.7, 55.6, 57.1, 58.3, 59.2, 59.8, and 60 °C) on amplification results were examined using the DNA samples of ‘Xingmeili’ (accession no. 1) and ‘Stanley’ (101) and optimized conditions, as described in Materials and Methods, were obtained. Under the optimized reaction system, 42 ISSR primers from University of British Columbia, Canada, were tested, among which 12 primers (Table 2) were selected for their reproducible and clear band patterns.

Table 2. ISSR primers used for the genetic diversity evaluation of 104 plum and related accessions and their amplification performance.

View Fullscreen

A total of 103 bands were scored with an average band number of 8.6 per primer across 104 accessions assayed. Among the 103 bands, 99 bands (96.1%) were polymorphic, and only four bands were monomorphic. These results indicated a high level of diversity in the 104 accessions under investigation.

Twenty-six accession-specific band patterns were identified (Table 3). In total, seven accession-specific band patterns were obtained for accession 23 (‘Wanhei’, a European cultivar with unknown name), three for accession 89 (‘Yingxiangbai’, a white-flesh local apricot cultivar), three for accession 96 (P. tomentosa, a common rootstock for plum in northern China), two for accessions 6 (‘Hongyeli’, red-leaf plum, a variant of P. cerasifera) and 101 (‘Stanley’, a U.S. hybrid-derived European plum cultivar), and one for accessions 2, 13, 16, 60, 99, 100, and 102, respectively.

Table 3. ISSR primers generating unique band patterns to some Prunus accessions tested.

View Fullscreen

There were two possible accession-specific bands among the total 103 bands obtained. One was Primer 853-450 bp for all cultivated European plum and another Primer 827-400 bp for ‘2-214’ (accession no. 60).

Genetic similarities between accessions were estimated using the Jaccard coefficient, and the similarity coefficient matrix was established (data not shown). The similarity values varied from 0.19 for ‘89-7-3’ (15) versus ‘Limixing’ (91) to 1.00 for ‘Xiangjiaoli’ (8) versus ‘Bianyan’ (9) and ‘Xinjiang #4’ (92) versus ‘Xinjiang #5’ (93). Apparently, ‘Xiangjiaoli’ (8) and ‘Bianyan’ (9) are duplicated accessions, as are ‘Xinjiang #4’ (92) and ‘Xinjiang #5′ (93).

‘Tianli’ (24) versus ‘Ganli’ (25); ‘Jinshuai’ (83) versus ‘Changli1 #5’ (84); and ‘Pingguoli’ (43) versus ‘Kuandiandali’ (45), which were all difficult to discriminate due to their morphological similarity, showed similarity coefficients of 0.97, 0.91, and 0.90, respectively. ‘Tianli’ (24) versus ‘Ganli’ (25) were both introduced European plum cultivars with unknown original names, and their Chinese translated names, “tianli” and “ganli,” mean “sweet plum” in Chinese. The similarity based on ISSR data suggested that the two accessions were possible synonyms.

Three accessions (8, 81, and 100) have the same name, ‘Xiangjiaoli’, and they belong to three different species, P. simonii, P. salicina, and P. ussuriensis, respectively. Similarity coefficients between them range from 0.56 for accession 8 versus accession 100 to 0.72 for accession 8 versus accession 81, correlating with the growth habit differences between the three mentioned accessions: P. simonii is upright, P. salicina is intermediate, and P. ussuriensis is spreading.

The similarity coefficients of four accession pairs with same names, ‘Xiangbianli’ (10) versus ‘Xiangbianli’ (11); ‘Bingtangli’ (12) versus ‘Bingtangli’ (13); ‘Ruth Grestetter’ (30) versus ‘Ruth Grestetter’ (53); and ‘Ozark Premier’ (54) versus ‘Ozark Premier’ (71) were 0.75, 0.76, 0.55, and 0.80, respectively, indicating that four pairs were all homonymous pairs incited by mislabeling.

Jaccard coefficients of similarity of ‘Lihe #1’ (48), ‘Lihe #2’ (50), and ‘Lihe #4’ (52) versus ‘Qiuli’ (46) were 0.96, 0.93, and 0.98, respectively, and these results supported the suggestion that the first three were probable bud mutants or seedlings of ‘Qiuli’ (Zhang and Zhou, 1998). The two pairs of improved plum cultivars, ‘Oishi wase’ (88) versus ‘Oishi nakade’ (104) and ‘Freedom’ (35) versus ‘Fortune’ (86) were sibling lines, but the similarity coefficients between them were only 0.66 and 0.64, respectively, showing that the similarity coefficients based on ISSR data did not coincide completely with their pedigree information.

The average Jaccard coefficients within and between species or groups (Table 4) indicated that similarities within species were higher than those between species. Genetic variability within P. salicina was lower than within P. domestica, and either five accessions of P. simonii or six accessions of wild European plum presented low genetic variability. The dendrogram using accessions grouped by species (Fig. 1) was constructed with a cophenetic correlation coefficient of 0.85 by Mantel test. The cophenetic correlation coefficient is considered to be a good representation of the data matrix in the dendrogram if it is ≥0.8 (Romesburg, 1990). This dendrogram showed that P. simonii was closer to P. salicina than was P. ussuriensis to P. salicina, although both were very similar to P. salicina in morphological traits. P. domestica was more closely related to P. spinosa than to P. cerasifera. Hybrids of plum and apricot were more similar to apricot than to plum.

Table 4. Average Jaccard coefficients of 10 Prunus species or groups within and between them based on ISSR data

View Fullscreen

View Figure

• Download as Powerpoint
• Download Figure

The clustering results including all 104 accessions tested (Fig. 2), with a cophenetic correlation coefficient of 0.96 by Mantel test, showed that they were divided into three main groups and outgroups. Group I was composed of all eight accessions of plumcot types and one apricot cultivar, Yinxiangbai (89). Group II consisted of all diploid plum accessions and three European cultivar types [i.e., ‘President’ (34), ‘Ruth Grestetter’ (53), and ‘2-214’ (60)]. This group has three subgroups, such as cherry plum (containing two accessions of P. cerasifera, 5 and 6), Chinese plum-types, and P. tomentosa (96). Group III included the tetraploid species [‘Heicili’ (99)] and all hexaploid European plum, including 6 accessions of wild European plum (17, 22, 29, 92, 93, and 94), except for three accessions of European plum including accessions 34, 53, and 60 which were mixed into Group II, ‘89-7-3’ (15) and ‘Wanhei’ (23) which were in the outgroups.

View Figure

• Download as Powerpoint
• Download Figure

Associations among 104 accessions were also revealed by PCoA (Fig. 3). In the PCoA plot, similar grouped results were obtained, but two European plum accessions, ‘89-7-3’ (15) and ‘Wanhei’ (23), which were divided into outgroups in the clustering result (Fig. 2), could be grouped within Group III, and additional information was revealed (e.g., the plumcots were displaced between apricot and Chinese plum-types that reflected their phylogenetic relationships), while Eurasian species (P. spinosa and P. cerasifera) between European plum and Chinese plum-types that reflected their geographic origins.

View Figure

• Download as Powerpoint
• Download Figure

Discussion

ISSR PCR using primers based on dinucleotide, tetranucleotide, or pentanucleotide repeats has now become fashionable among researchers (Zietkiewicz et al., 1994). ISSRs have been used in fruit trees (Archak et al., 2003; Cekic et al., 2001; Fang et al., 1997; Goulao et al., 2001; Levi and Rowland, 1997; Moreno et al., 1998; Sankar and Moore, 2001), but little information is available for evaluation and identification of plum resources using ISSR techniques. Our study further supports ISSR as a reliable, quick, uncomplicated, and cheap method of identification for the large number of accessions held in most collections (Gilbert et al., 1999). This method could be used to eliminate duplicated collections, to discriminate among homonyms or synonyms, to identify closely related accessions, to measure genetic diversity, and to reveal genetic relationships.

Zietkiewicz et al. (1994) noted that primers anchored at the 5′ end displayed more specificity than those anchored at the 3′ end. Fang et al. (1997) also found that in citrus, 5′-anchored primers generally had more specificity and thus generated fewer but larger fragments than 3′-anchored ones. However, 10 of 12 primers used in this study were anchored at the 3′ end, and they produced clear patterns too, which is in agreement with the results obtained by Huang and Sun (2000). Wang et al. (1994) reported that (AT) _n was the most abundant microsatellite in plant nuclear genomes, but in the present study the primer UBC 880 with the motif of (GGAGA)₃ generated maximum fragments, suggesting that this pentanucleotide repeat was most frequent among eight motifs tested in plum genomes.

Regarding the effects of primers with same motifs but differently anchored nucleotides, primers UBC 827, 855, and 857 amplified 4, 6, and 12 bands, respectively, and this result supported the observation by Ammiraju et al. (2001) that anchored nucleotides at the 5′ and 3′ ends of the primers play important roles in detecting polymorphism.

Somewhat unexpectedly, the European plum accessions—‘President’ (34), ‘Ruth Grestetter’ (53), and ‘2-214’ (60)—were clustered into same group as diploid accessions, mainly Chinese plum-types. Similar overlaps were observed in the plum dendrogram generated with RAPD characters by Shimada et al. (1999), where three Chinese plum-types, namely ‘King’, ‘Jinsantou’, and ‘Beauty’, were classified into the “European plum group”; these results were considered plausible if these cultivars are interspecific hybrids between Japanese plum (P. salicina) and related species such as P. cerasifera or P. simonii.

In general, the phenogram showed clustering in good accordance with the classification previously established. One apricot accession (89) was clustered into the same group as all plumcot types, suggesting that seven accessions of hybrids between plum and apricot were possibly more closely related to apricot than to plum. Possible explanations for this were 1) repeated back-crossing plum–apricot hybrids with apricots or 2) favorable selection of the apricot-like descendant from segregating populations as the main objective for the hybridization between plum and apricot in apricot breeding programs as necessary to select a new apricot cultivar with a late flowering time so as to elevate the resistance to spring frost injury for apricot, which has been a serious problem for apricot industry worldwide.

Group II, a diploid species group, included all accessions of P. salicina, P. cerasifera, P. simonii, P. ussuriensis, P. tomentosa, and P. salicina hybrids, among which the first five species were all of Asian origin and the last one was interspecific hybrid between P. salicina and other diploid species. Three subgroups were formed in this group, such as the cherry plum subgroup including ‘Yingtaoli’ (5) and ‘Hongyeli’ (6), the Chinese plum subgroup including P. simonii, P. salicina, P. ussuriensis, and three accessions of P. domestica (accessions 34, 53, and 60), and P. tomentosa subgroup (96).

Chinese plum, P. salicina, was native to China, and introduced into Japan no earlier than 1500 ad. Many cultivars were introduced from Japan into California around 1870 and subsequently to Europe. Luther Burbank introduced several Japanese plum cultivars and hybridized them with American native species, resulting in a series of improved plum cultivars, which laid a foundation for fresh market plum production worldwide (Ryugo, 1988). Five plum cultivars from Japan, including ‘Akihime’ (39), ‘Taiyo’ (74), ‘Gecko’ (77), ‘Oishi wase’ (88), and ‘Oishi nakade’ (104), were distributed across the Chinese plum subgroup in present dendrogram, indicating that diverse plum resources had been introduced into Japan from China. All the improved Chinese plum hybrids, although most of them derived from interspecific hybridization, were grouped together with pure Chinese plum cultivars possibly due to back-crossing or selection bias, and this result was in agreement with the P. salicina-predominating genetic components of improved Chinese plum hybrids (Byrne and Littleton, 1988).

The clustering results did not reflect the geographic origin for both Chinese plum-types in Group II and European plum in Group III (Fig. 2), but the plumcot types were plotted between apricot and Chinese plum and Eurasian species (P. spinosa and P. cerasifera) between P. domestica and Chinese plum (P. salicina and P. simonii) in the PCoA plot (Fig. 3), which was consistent with their genetic origin and geographic distribution, respectively.

Two European cultivars, ‘89-7-3’ (15) and ‘Wanhei’ (23), were grouped into distinct outgroups, and another three cultivars [i.e., ‘President’ (34), ‘Ruth Grestetter’ (53), and ‘2-214’ (60)] were intermixed with diploid accessions, suggesting that they may be useful materials to broaden the genetic background of European plum cultivars.

As to the taxonomic placement of P. simonii, three opinions exist: the first one is that P. simonii is derived from hybridization between apricot and plum (Faust, 1999; Luo et al., 1992; Ruan et al., 2002); the second is that it is thought as an independent species (Qu and Sun, 1990; Wu, 1984; Yu, 1984; Zhang and Zhou, 1998); and the last opinion considers P. simonii to be a variant of P. salicina (Okie and Weinberger, 1996; Yoshida, 1987). Our results support P. simonii as a variant of P. salicina because all five accessions of P. simonii formed a subcluster underneath Group II and were inserted among Chinese plum-types, coinciding with the clustering results with RAPD markers (Boonprakob and Byrne, 2001; Liu et al., 2006).

Prunus salicina was native to Yangtze River basin of China, while P. simonii was native to northern China. The present study implied that one or a few individuals of P. salicina from Yangtze River basin colonized in a vacant habitat of northern China and a variant was formed as the gene pool of P. simonii had low levels of genetic variation (Table 4). In other words, the “founder effect” may have played an important role in the speciation of P. simonii.

Although most Chinese taxonomists thought P. ussuriensis to be an independent species, Faust (1999) suggested that it was only a minor variation of P. salicina. Compared with P. simonii, P. ussuriensis was distinct from P. salicina and its hybrids, so it seems reasonable to consider P. ussuriensis as an independent species.

The discovery of putative wild P. domestica in a canyon in the Yili region in Xinyuan County, Xinjiang, China (Lin and Shi, 1990), suggested the possibility that European plum may have originated in China. The plants found in Xinjiang, according to their morphological characteristics (Zhang et al., 1998) and chromosome numbers (Lin et al., 1991), belong to P. domestica. In present study, six accessions collected from the Tiansan Germplasm Repository for Wild Fruit Resources were clustered into a subcluster underneath Group III, the European plum group, further confirming that they belong to P. domestica. High average Jaccard coefficients were presented for six accessions of wild P. domestica from six populations from Xinjiang, China (Table 4). This result implied that the cultivated P. domestica was brought to Xinjiang and had naturalized there and that Xinjiang was not the origin center of P. domestica, as a high level of diversity should exist in the center of origin of a wild species (Vavilov, 1930). A similar result was obtained by us with RAPD markers (Liu et al., 2006).

Our results suggest that P. domestica was derived from P. spinosa because both were closely related. The limitation of our experiment is that only one accession of P. spinosa is used, so collecting more accessions of related species and further study on the origin of European plum based on DNA techniques is needed to resolve this question. The evaluation of genetic relationships between European plum-related species is presently under way using cpDNA and mtDNA PCR-RFLP methods in our laboratory.

Classification of stone fruit species—including plum, apricot, cherry, and peach—has long been complicated and problematic due to the diverse phenotypes and large number of species. Some Chinese pomologists still put plum, apricot, cherry, and peach into separate genera, as Prunus L., Armeniaca Mill., Cerasus Mill., and Amygdalus L., respectively (Yu, 1984), and refer to the plum genus as the “small Prunus,” but the most widely adopted classification of stone fruit species places all of them under one genus (i.e., Prunus), the so-called “large Prunus” (Lee and Wen, 2001). Our results support the classification of stone fruit species under “large Prunus” because apricot (89), plumcot types (1, 2, 3, 4, 90, 91, 97, and 98), and P. tomentosa [named Cerasus tomentosa (Thunb.) Wall. by “small Prunus” taxonomists] (96) were grouped with the diploid plum species group earlier than European plum (Fig. 2).