Scientia Horticulturae 70 (1997) 67-72

Comparative study of Chinese tree peony cultivars by random amplified polymorphic DNA (RAPD) analysis

T. Hosoki ^a*, D. Kimura ^a, R. Hasegawa ^a, T. Nagasako ^a, K. Nishimoto ^a, K. Ohta ^a, M. Sugiyama ^b, K. Haruki^b

^a Department of Agriculture, Shimane University, Nishikawatsu-cho 1060. Matsue. Shimane 690, Japan

^b Shimane Agricultural Experiment Station, Ashiwata-cho 2440, Izumo. Shimane 693. Japan

Received 7 October 1996

Corresponding author Fax: 00-81-852-32-6-6499

Random amplified polymorphic DNA (RAPD) analysis was applied to Chinese tree peony cultivars (Paeonia suffruticusa var. spontanea) to clarify their genetic relationships. Of 40 decamer primers tested. 11 produced 92 useful polymorphic DNA bands. Using these bands, 19 cultivars were distinguished and the similarity values among them were calculated. Although the values between cultivars were high. ranging from 0.722 to 0.928. a dendrogram produced by cluster analysis revealed that Chinese cultivars were largely divisible into four groups. These groups did not necessarily coincide with those classified based on flower and leaf types or petal color.

© 1997 Elsevier Science B.V.

Keywords: Paeonia sitffruticosa var. spontanea; Random amplified polymorphic DNA: Cultivar classification

1. Introduction

Chinese tree peony (Paeonia suffruticosa var. spontanea Rehd.) was developed as an ornamental plant from medicinal plants in the 5th century (Yu. 1980). Thereafter, many cultivars have been bred over some 1500 years. Cultivars bred during the 10th to 13th centuries are still growing nowadays along with other cultivars bred from the 17th to 20th centuries. Chinese cultivars have unique traits which are rarely seen in Japanese cultivars (P. suffruticosa Andr.) or in American/French cultivars (Paeonia lutea Delav. ex Franch. or Paeonia delavayi Franch. X P. suffruticosa). Many of the Chinese cultivars bear showy, heavy, double or proliferated flowers. They flower in late April earlier than the Japanese or American/French cultivars, which flower in early to middle May (Hashida, 1990). Furthermore, Chinese cultivars are relatively small (Liu et al 1987) and are considered as potential pot plants as well as garden plants (Hosoki and Kimura, 1996). Popular Chinese cultivars have been classified based on flower and leaf type (Yu, 1980) and petal color as well as anthocyanins (Hosoki et al., 1991). Although this classification gives useful information for cultivar breeding, it is insufficient to clarify genetic distances/relationships among cultivars or to distinguish them accurately.

Recently, direct DNA analysis by random amplified polymorphic DNA (RAPD) has been shown to be effective in clarifying species/cultivar genetic relationships in ornamental trees such as azalea (Kobayashi et al., 1995) and rhododendron (Iqbal el wr 1995). In Paeonia spp., Pei (1993) studied DNA of wild plants (P. suffruticosa var. spontanea and var. papaveeracea) grown in the mountains of China by polymerase chain reaction (PCR) RAPD and found genetic differences between plants. More recently, Sang et al. (1995) clarified reticulate evolution of herbaceous peonies by determining nucleotide sequence of nuclear ribosomal DNA. These studies deal with genetic differences between species in tree or herbaceous peonies. Thus, no studies on horticultural cultivars of the Chinese tree peony ( P. suffruticosa var. spontanea) by PCR-RAPD have been reported while those on Japanese tree peony cultivars (P. suffruticosa) were recently investigated (Hosoki et al., 1997).

The objective of this experiment was to clarify the genetic relationships of Chinese tree peony cultivars and to discriminate them using RAPD by PCR. Since Chinese tree peony cultivars have been vegetatively propagated by division (Yu. 1980), genetic traits of individual cultivars have been basically retained.

2. Materials and methods

2. 1. Plant materials

Nineteen Chinese tree cultivars (Paeonia suffruticosa var. spontanea) grown in Tree Peony Collection of Yatsuka-cho. Shimane Prefecture. Japan, were tested in this study. Their flower form, petal color, and the date and location of collection are shown in Table 1. Fresh, newly expanded leaves from each cultivar were harvested in late April 1995 and stored at -20°C until DNA extraction.

2.2. DNA extraction

Stored leaves (100 mg) were shredded into pieces of about 1 mm width and DNA was extracted using a non-grind DNA extraction kit (Isoplant. Nippon Gene Co. Ltd., Osaka, Japan), according to the manufacturer's instructions. The extract was dissolved in 50 micro- l of TE buffer containing 10 mM Tris-HCl and 1 mM EDTA at pH 8.0. Since DNA extracted by this method does not always produce stable amplification products by PCR. the DNA extract in TE buffer was further purified by the CTAB protocol (Stewart and Via. 1993). Repeating this purification procedure four times produced stable PCR products; purified DNA free of contaminants showed OD₂₆₀/OD₂₈₀~ 2.1 when assessed spectrophotometrically.

2.3. DNA amplification

Amplification was achieved by the method of Sugiyama et al. (1995) with slight modification. Forty decamer primers (oligonucleotides) (Kits A and B, Operon Technologies. Alamanda, CA, USA) were tested, of which 11 primers were selected for RAPD analysis of Chinese tree peony cultivars. Amplification reactions were conducted in 25 micro-l volumes containing 10 mM Tris-HCl (pH 8.3), 50 ml MgCl₂;. 0.2 mM d-NTP. 2 micro-M primer. 0.05 units micro-l^-1 Taq DNA polymerase (Takara. Co. Ltd.. Ohtsu, Shiga. Japan) and 0.7 micro-g template DNA dissolved in 1 micro-l TE. Each reaction tube was overlaid with mineral oil to prevent evaporation. Reactions were carried out in a Perkin-Elmer Ceius Gene Amp PCR system 9600 (Poster, CA, USA) programmed as follows: 3 min at 94 C for initial strand separation, 50 cycles of 45 s at 94°C. 1 min at 40°C and 1 min 3t 70-C. followed by 5 min at 70°C for final extension. PCR products were stored at 4°C before analysis. Amplification products were analyzed by gel electrophoresis in 1.5% agarose gel in TAE buffer containing 0.04 M Tris-acetate and 0.001 M EDTA. stained with ethidium bromide, and photographed under UV light. Approximate molecular sizes of amplification products were estimated using a DNA size marker, lambda/Hind II digest- phi-chi 174/Hinc II digest (Toyobo Co., Tokyo. Japan). All reactions were repeated three times and only reproducible bands were adopted for data analysis.

2.4. Data analysis

Amplified products were analyzed by pairwise comparisons of cultivars based on the percentage of common fragments, and a similarity matrix was generated (Nei and l.i, 1979). A dendrogram was constructed using the similarity matrix data by applying the unweighted pair group method with arithmetic average (UPGMA) cluster analysis (SYSTAT program of Macintosh. Apple Computer Inc.. Cupertino. CA. USA).

3. Results and discussion

Of 40 primers, 11 produced 9-14 DNA bands per primer suitable for data analysis of the 19 cultivars. Those primers which produced either too few or more than 15 bands were not considered useful because they were monomorphic or too difficult to distinguish from comparable bands of cultivars on the agarose gel. A total of 130 bands were produced from the 11 useful primers, ranging in size from 300 to 1540 base pairs (Table 2) (Fig. 1). The 19 cultivars were classified using the 92 polymorphic bands obtained from the 11 primers. As seen in the dendrogram (Fig. 2). Chinese cultivars were largely separated at similarity values 0.826. 0.307. 0.816. and 0.828 into four groups, (a: 8. 10. 14). (b: 3. 5. 11. 12. 13. 16). (c; 2. 6. 7, 9, 15. 17. 18. 19) and (d: 1. 4). respectively. although overall similarity values between the cultivars were relatively high.

Yu (1980) classified 80 Chinese cultivars into 17 groups based on nine flower types and, two leaf types. The grouping obtained using PCR-RAPD was compared with his classification. 'Shou an hong' (1) and 'Yao huan' (4) in group d agreed with his classification, in which both cultivars belonged to crown-flower with incurved leaf type. 'Wo rong bang shen' (9) and 'Jin yu jiao zhang' (19) in group c belonged to headed bomb-flower type. but the former belonged to incurved leaf type and the latter, to flat leaf type. 'Lang tien Yu’ (12) and 'Qing long wo mo chi' (13) in group b belonged to anemone-flower with incurved leaf type, and crown-flower w ith flat leaf type, respectively, in his classification. Thus, cultivar grouping by PCR-RAPD did not completely agree with Yu's classification based on flower and leaf type. Petal colors shown in Table I and petal anthocyanins (pelargonidin, cyanidin and peonidin) (Hosoki et al., 1991) were not related to cultivar position in the dendrogram. Mos probably, the genes associated with horticulturally evaluated traits were limited and were not relevant to genes which were amplified by the present random primers. Further study to find DNA markers would be necessary if a specific trait with horticultural value was to be targeted in a breeding program of tree peony.

Similarity values between the remotest cultivars and the nearest cultivars were 0,722 and 0.891, respectively. In contrast, the same similarity values in Japanese tree peony cultivars were 0.807 and 0.928, respectively, by the same calculation method (Hosoki et al., 1997). Although for many Chinese cultivars the date and location of collection is not known, relatively lower similarity values between them might reflect the long time period (10th to 20th century) and different locations in which Chinese cultivars have been bred. Most of the Japanese cultivars grown nowadays were bred in the 19th and 20th centuries and all of them originated from one location. Osaka Prefecture (Hashida. 1983). Thus, Chinese cultivars bred at different times and in differing locations are still being cultivated with relatively large genetic diversity, suggesting that the potential fur creating new cultivars is high.

In addition, in a breeding program for creating cultivars with hybrid vigor, a dendrogram would be useful to select parents which are genetically as remote as possible, e.g. 'Zhao feng' (10) and 'Yao huang' (4) (Fig. 2).

In conclusion, PCR-RAPD analysis of Chinese tree peony would offer additional information for a breeding program of Chinese cultivars which, until now, has been planned based solely on observational traits of plant and flower morphology and petal color.