Plant materials.

We sampled 40 accessions from the Xiongyue Arboretum (Yingkou, China) of which five accessions (three species) belonged to subgenus Prunophora, five accessions (three species) to subgenus Amygdalus, eight accessions (four species) to subgenus Lithocerasus, and 22 accessions (seven species and two hybrids) to subgenus Cerasus (Table 1).

Table 1.

Forty Prunus accessions used in genetic relationship analysis among 10 rootstock species from China by simple sequence repeat.

View Table

Genomic DNA extraction.

Total genomic DNA was extracted from young expanded terminal leaves using the citrimonium bromide method (Doyle and Doyle, 1990) with minor modifications. Then, the concentration and quality of the DNA were assessed using ultraviolet–visible spectrophotometry (DU 800; Beckman Coulter, Fullerton, CA) and electrophoresis on 1% TAE agarose gels, and the samples were diluted to a final concentration of 250 ng·μL^−l.

PCR amplification and product electrophoresis.

The diluted genomic DNA was PCR amplified using 24 sets of SSR primers that were previously developed in peach, apricot, and cherry (Table 2). The 20-μL reaction mixture contained 20 mm Tris-HCl (pH 8.4), 50 mm KCl, 4 mm MgCl₂, 0.1 mm each dNTP, 0.2 M each primer, 40 ng genomic DNA, and 0.45 U Taq polymerase and were amplified using a thermocycler (PTC-200; MJ Research, Waltham, MA) with the following program: an initial denaturation of 4 min at 94 °C; 28–31 cycles of 45 s at 94 °C, 45 s at 57.5–62 °C, and 45 s at 72 °C; and a final extension of 8 min at 72 °C. The amplification products were electrophoresed on 8% polyacrylamide gels in 1× TBE buffer with a constant current of 40 mA, and the gels were stained with silver nitrate (Promega, Madison, WI), according to the manufacturer’s instructions. In addition, amplification with each primer combination was performed twice to verify the reproducibility of our results.

Table 2.

List of simple sequence repeat primers used in genetic relationship analysis and their amplified products in 40 Prunus accessions.

View Table

SSR data analysis.

Each SSR band was considered an individual character, and the presence or absence of the band was scored using binary code (present = 1, absent = 0). Subsequently, a data matrix was assembled and analyzed using PAUP* (Swofford, 1998), and a pairwise distance matrix was generated based on total character differences. The genetic relatedness of the Prunus species was analyzed using the unweighted pair group method with arithmetic mean (UPGMA), based on total character difference, and bootstrap analysis was conducted using UPGMA search methods with 10,000 replicates, to obtain confidence values for the branches of the tree.

SSR marker characterization.

Bands or banding patterns that are unique to specific groups can be used for species identification or subgenus classification (Zhang et al., 2008). To identify the more common alleles shared by different species, 24 SSR primer sets were selected from over 170 pairs of primers that were previously developed for peach, cherry, and apricot (Zhang et al., 2008). In the present study, a total of 251 alleles were identified, with a mean of 10.5 alleles per SSR locus. Among the 24 loci, the CPPCT27 locus was the most polymorphic, with 17 alleles, whereas the UCD-CH12 locus was the least polymorphic, with only six alleles (Table 2). Both common and unique alleles were identified using 10 group-specific markers on a certain set of species (Table 3).

Table 3.

Ten group-specific simple sequence repeat markers applied to genetic relationship analysis among 40 Prunus accessions distinguish a set of species by generating common and unique bands.

View Table

For example, one allele at the UDP96-003 locus was shared by accessions of P. tomentosa, P. humilis, P. japonica, and P. glandulosa and, therefore, was able to distinguish species in subgenus Lithocerasus from those in the other subgenera, whereas another allele from the same locus was shared by P. triloba and P. salicina (Fig. 1A). Similarly, the PacA10 locus was able to distinguish P. salicina (subgenus Prunophora) from P. tomentosa, P. triloba, P. humilis, P. japonica, P. glandulosa (subgenus Lithocerasus), P. davidiana, and P. persica [subgenus Amygdalus (Fig. 1B)]; and the UCD-CH31 and UCD-CH39 loci were able to distinguish P. tomentosa from all other species. In addition, both the UDP96-008 and UDP97-401 loci yielded species-specific alleles for each of P. japonica, P. humilis, and P. glandulosa; and the UDP96-008, UDP98-024, and CPPCT27 loci yielded a common allele for P. davidiana and P. persica (subgenus Amygdalus) (Table 3).

View Full Size

Simple sequence repeat band patterns of 21 Prunus accessions generated by UDP96–003 and PacA10 markers. (A) Marker UDP96–003: allele “a” was shared by Prunus tomentosa, Prunus humilis, Prunus japonica, and Prunus glandulosa; allele “b” was shared by Prunus triloba and Prunus salicina. (B) Marker PacA10: allele “c” was shared by Prunus armeniaca and Prunus sibirica; allele “d” was shared by P. tomentosa, P. triloba, P. humilis, P. japonica, P. glandulosa, Prunus davidiana, and Prunus persica. Prunus accession numbers are detailed in Table 1; M = DNA size markers.

Citation: Journal of the American Society for Horticultural Science J. Amer. Soc. Hort. Sci. 141, 5; 10.21273/JASHS03827-16

• Download Figure
• Download figure as PowerPoint slide

In addition, we also found that subgenus Cerasus (P. avium, Prunus fruticosa, Prunus cerasus, P. pseudocerasus, P. serrulata, Prunus mahaleb, Prunus yedoensis, P. cerasus × Prunus canescens, and P. avium × P. pseudocerasus) could be distinguished by a unique allele at the Pchgms2 locus. Although subgenus Cerasus exhibits a wide geographic distribution, all the specimens included in the present study possessed the same allele, whereas the other species exhibited polymorphism. During the optimization of Pchgms2 amplification, only a single allele was amplified under different annealing temperatures (52, 55, 58, and 60 °C) in the 22 accessions representing subgenus Cerasus. Thus, the locus Pchgms2 was very useful for identifying subgenus Cerasus.

Among the 24 loci, the UCD-CH12 locus was determined to be the most suitable for homology analysis, with four common alleles shared by the 17 species and 2 hybrids analyzed, and could be useful for elucidating relationships within the genus. Thus, it might eventually be possible to classify Prunus species using only one or two SSR markers. Cipriani et al. (1999) reported that 59% of peach SSR primer pairs amplified products of expected size in other Prunus species. Our results also show that the examined Prunus species were closely related and that SSR primer pairs exhibit high transferability within the genus. Although Dirlewanger et al. (2002) reported that the transferability of primer pairs across species is not related to genetic distance, our results indicated that more closely related species were more likely to share similar alleles. For example, P. humilis, P. japonica, and P. glandulosa shared a distinguishing allele with P. salicina at locus UDP98-416; and P. humilis, P. japonica, P. glandulosa, P. triloba, and P. tomentosa shared a distinguishing allele with P. davidiana and P. persica at locus PacA10 (Table 3), which indicated that subgenus Lithocerasus was more closely related to subgenus Prunophora than subgenus Cerasus.

Phylogenetic analysis.

All the accessions from subgenus Cerasus clustered together (Fig. 2), thus providing support for their common genetic origin. P. cerasus and P. fruticosa were more closely related to each other than to P. avium, which supports the hypothesis that P. cerasus arose as an allotetraploid from the hybridization of P. avium (2n = 2x = 16) and P. fruticosa (2n = 4x = 32; Olden and Nybom, 1968). Meanwhile, the hybrid ‘Gisela 5’ (P. cerasus × P. canescens) connected the P. avium group to the group that contained P. cerasus and P. fruticosa, and the hybrid ‘Colt’ (P. avium × P. pseudocerasus) was grouped between the cherries from Asia and Europe. Although P. mahaleb originated in Europe, it was closely related to cherries from Asia. Generally, subgenus Cerasus is more divergent from the other groups. Subgenus Cerasus was also more closely related to subgenus Prunophora than to subgenus Amygdalus, which confirms the hypothesis that cherry is genetically more distant from peach than from apricot (Badenes and Parfitt, 1995; Chin et al., 2014; Watkins, 1976). Successful hybridization among Prunus species has been reported [e.g., Prunus amygdalus × P. persica (Felipe, 2009), Prunus domestica × P. armeniaca (Jones, 1928), P. davidiana × Prunus mira (Shoferistov, 1989)], which indicates that hybridization and crossing could occur among plum, apricot, and peach, but not with cherry.

View Full Size

Dendrogram of 40 Prunus accessions based on unweighted pair group method with arithmetic mean analysis of simple sequence repeat data. Numbers below the lines indicate bootstrap values (percentage of 1000 replicates). Bootstrap values greater than 50% are shown.

Citation: Journal of the American Society for Horticultural Science J. Amer. Soc. Hort. Sci. 141, 5; 10.21273/JASHS03827-16

• Download Figure
• Download figure as PowerPoint slide

The P. salicina accessions formed a group with high bootstrap values, along with a group that included P. humilis, P. japonica, and P. glandulosa (Fig. 2), which together formed a group with P. tomentosa and P. triloba. The apricot group (P. armeniaca, P. sibirica) was more similar to the plum group (P. salicina) than to the peach group (P. davidiana, P. persica). A previous morphological study suggested that the evolution of these species progressed from peach > apricot > plum > cherry (Yü and Li, 1986), which is corroborated by the molecular analysis of the present study.

Genetic diversity of four species from subgenus Lithocerasus.

The four species from subgenus Lithocerasus that were included in the present study failed to cluster into a single group and, instead, were grouped with members of the subgenus Prunophora. This grouping suggests that subgenus Lithocerasus is closely related to P. salicina. A close relationship was previously reported between P. glandulosa and P. japonica, which belong to subgenus Prunophora and subgenus Amygdalus, respectively, based on RAPD analysis (Shimada et al., 2001). The present study grouped P. japonica, P. glandulosa, and P. humilis, and within this group, P. glandulosa was more closely related to P. japonica than to P. humilis. In addition, a previous isozyme study (Mowrey and Werner, 1990) indicated that P. japonica only differed from P. glandulosa by one of 110 alleles and, moreover, that P. japonica and P. glandulosa grouped together in subgenus Cerasus. However, although the present study confirms the genetic similarity of P. humilis, P. japonica, and P. glandulosa, the species were placed in subgenus Prunophora (Fig. 2), in agreement with the ITS sequence analysis of Bortiri et al. (2006), which indicated that P. tomentosa and P. glandulosa were more closely related to subgenus Prunophora than to subgenus Amygdalus.

Evidence from interspecific crosses also supports the classification of subgenus Lithocerasus as distinct from subgenus Cerasus and suggests that subgenus Lithocerasus is more closely related to subgenus Prunophora. For example, members of subgenus Lithocerasus section Microcerasus have been reported to hybridize readily with members of subgenus Prunophora and, to some extent, with members of subgenus Amygdalus, but generally fail to hybridize with members of subgenus Cerasus (Garley, 1980; Kataoka et al., 1988). Furthermore, P. japonica is also capable of growing on rootstocks of both P. glandulosa and P. armeniaca (Fan, 1999), and the grafting compatibilities between some members of subgenus Lithocerasus and subgenera Prunophora or Amygdalus is greater than that of subgenus Cerasus (Ding, 1997; Mizutani et al., 1985). Thus, our results support a classification system that separates bush cherry and rock cherry into either subgenus Lithocerasus or subgenus Prunophora.

The status and relationship of P. tomentosa and P. triloba.

The species P. tomentosa and P. triloba are morphologically similar and are mainly distinguished by fruit morphology. The fruit of P. triloba is larger and dry at maturity, whereas the fruit of P. tomentosa is smaller and remains fleshy. However, it is difficult to distinguish the two species without flowers or fruit, and few studies have investigated their phylogenetic relationship. The topology of the present study agrees with the species’ morphological similarity, since P. tomentosa and P. triloba were grouped together and formed a sister group to subgenus Prunophora. Previously, RAPD analysis (Shimada et al., 2001) placed the two species in subgenus Amygdalus; however, isozyme analysis grouped the species together as a separate clade that was closer to members of subgenus Prunophora than to those of subgenus Cerasus (Mowrey and Werner, 1990). Further, according to the analysis of nrITS sequences and plastid markers (Chin et al., 2014), P. tomentosa and P. triloba are grouped together and are more closely related to P. glandulosa (subgenus Lithocerasus) and subgenus Prunophora than to subgenus Amygdalus or subgenus Cerasus. Our findings support the conclusions drawn from the analysis of isozymes, nrITS sequences, and plastid markers.

Traditionally, P. triloba has been placed in subgenus Amygdalus; however, this classification has been questioned. For example, Zhao (1996) suggested that P. triloba be placed in subgenus Cerasus, along with P. tomentosa, based on morphology. In addition, although P. triloba can be used as a rootstock for plum, apricot, and peach, it causes dwarfing in peach (Ding, 1997), a phenomenon that might be explained by genetic distance, since P. triloba is more closely related to subgenus Prunophora (plum and apricot) than to subgenus Amygdalus (peach, nectarine, and almond). Our results also support this point of view.

Meanwhile, P. tomentosa cannot be used directly as a rootstock for sweet cherry, owing to the incompatibility of the two species, although using P. tomentosa as a rootstock and Fuli (P. salicina) as an interstock produces sweet cherries that grow well and fruit earlier. Grafting compatibility also suggests that P. tomentosa is more closely related to subgenus Prunophora than to subgenus Cerasus, since a hybrid rootstock for peach was derived from a cross between P. tomentosa and Prunus cerasifera (Warner, 1998). The interspecific hybridization of P. tomentosa and P. salicina (Zhang and Gu, 2015) also supports the conclusion that P. tomentosa is more closely related to subgenus Prunophora than to subgenus Cerasus. Therefore, the results of the present study suggest that P. triloba and P. tomentosa should be assigned to the same group, either in subgenus Lithocerasus or subgenus Prunophora, rather than in subgenus Amygdalus.

Implications for Prunus rootstock breeding.

The genetic resources of Chinese Prunus species have made important contributions to global fruit production, either by direct planting or as breeding material. In addition, Chinese Prunus species may also be useful as a source of genetic variation for breeding Prunus rootstocks, especially since interspecific crosses in the genus are usually feasible and efficient (in terms of selection pressure). Such interspecific crosses permit the introgression of beneficial genes for tolerance to drought and cold and resistance to pests and disease, from wild species to cultivated species, which can result in progeny that can be cultivated over a wide range of climates or have a wide range of compatibility with other cultivars.

To identify suitable rootstocks for stone fruits, grafting experiments have typically focused on more primitive species from China, such as P. pseudocerasus, P. canescens, P. tomentosa, P. humilis, P. japonica, and P. triloba, which are characterized by their cold resistance, dwarfing, and early fruiting. Further, distant hybridization within Prunus species has been widely applied as the preferred strategy for fruit production and has yielded a number of cultivars. For example, ‘Colt’ (P. avium × P. pseudocerasus) and ‘Gisela 5’ and ‘Gisela 6’ (P. cerasus × P. canescens) have been important rootstocks for sweet cherry (Bielicki and Rozpara, 2010), whereas a hybrid rootstock from a cross between P. tomentosa and P. cerasifera is used for peach (Warner, 1998), and an interspecific hybrid of P. tomentosa and P. salicina exhibited dwarfing, large fruit, and cold resistance (Zhang and Gu, 2015). However, an improved understanding of the genetics of Prunus species is fundamental to the development of new breeding programs, and the close relationships identified in the present study will improve Prunus breeding programs by helping breeders choose parental combinations and allow the combination of some major characteristics from wild species.