The Korean Tea Plant (Camellia Sinensis): RFLP Analysis of Genetic Diversity and Relationship to Japanese Tea

Breeding Science 54 : 231-237 (2004) The Korean Tea Plant (Camellia sinensis): RFLP Analysis of Genetic Diversity and Relationship to Japanese Tea

Satoru Matsumoto*1), Yoshikazu Kiriiwa2) and Satoshi Yamaguchi3)

1) National Institute of Vegetable and Tea Science, 360 Kusawa, Ano, Mie 514-2392, Japan 2) Faculty of Agriculture, University of Shizuoka, 836 Ohya, Shizuoka 422-8529, Japan 3) Faculty of Agriculture, University of Ehime, 3-5-7 Tarumi, Matsuyama 790-8566, Japan

We used RFLP analysis with phenylalanine ammonia- extending from south-east China to Assam in north-east lyase (PAL) cDNA as a probe to evaluate the genetic India (Eden 1958). On the origins of Korean tea plants, opin- diversity of the Korean tea plant (Camellia sinensis ions vary; some believe that it was introduced from China in var. sinensis). We analyzed 297 plants collected from A.D. 828 and the seeds were sown on Mt. Jiri; others consid- the grounds of 6 old temples and from 1 tea farm. The er that tea was already present in Korea at the beginning of patterns of DNA fragments in plants from the 6 temples the 6th century (Kimura 1993, Choi 2000, Chun 2000). His- were variable and differed from those of Japanese teas. torically, tea culture was tightly associated with Buddhism In Japanese teas the PAL locus is composed of 3 multi- in ancient Korea; it seems that Buddhist monks from China fragment alleles, but at least 10 fragment alleles were took tea seeds and sowed them around their temples. Even apparent in the Korean teas. The Korean teas showed now, the tea garden remains part of the typical Korean tem- greater genetic diversity than Japanese teas. The RFLP ple. It is considered that tea planting spread from such tem- patterns of all 12 samples taken from the tea farm, ples, and tea farms were then developed. Further, tea seeds where Japanese tea seeds had been introduced histori- were introduced in large quantities from Taiwan and Japan cally, were the same as those of the Japanese teas. The in the early 20th century, when Japanese immigrants to Korean teas were divided into 2 different genetic Korea began to establish tea farms. Therefore, tea was intro- groups; one group was found around old temples and duced from China many centuries ago and from Taiwan and was probably derived from China. The other originated Japan in recent times. from Japanese teas that were introduced 50 to 100 years Although there has been a culture of tea drinking for ago. RFLP analysis using PAL cDNA was very useful 2000 or more years in Korea, the development of the tea infor the detection of genetic diversity in Korean teas, be- dustry was slow, and production was small. One of the rea- cause the results of this analysis were similar to those of sons was that many generations of breeding, together with previous RAPD and morphological studies and were natural selection, were required to adapt tea cultivars to the able to reveal the existence of the 2 tea groups. Our re- severe winters of Korea. However, suitable tea cultivars sults showed that not only several morphological char- have now been developed, and the amount of cultivation is acters, but also the genetic background of Korean teas, increasing. Breeding of new cultivars suited to the Korean differed from those of Japanese teas. Although further climate and soil conditions is also being conducted by the evaluation of Korean teas as genetic resources is re- National Research Institute of Tea in Korea (Choi 2000). quired, it is possible that Korean teas will prove useful Some characteristics of Korean teas are different from in Japanese tea breeding. those of Japanese teas, for example, the cold resistance of Korean tea plants is superior. Analysis of Korean teas can be Key Words: Camellia sinensis, Korean tea, Phenyl- used as a tool for judging whether they can be used as breed- alanine ammonia-lyase (PAL), RFLP, ing materials for Japanese green teas. Some researchers have Genetic diversity. evaluated the morphological characteristics (Ikeda and Park 2002), cold hardiness, disease resistance (Takeda 2002) and DNA diversity of Korean teas by using RAPD markers (Lee et al. 1995, Kaundun et al. 2000). However, more informa- Introduction tion is needed to clarify the diversity of Korean teas and how they differ from Japanese teas. Tea, Camellia sinensis, is an important beverage crop Tea leaf contains catechins, which are synthesized that is indigenous to the Asian region. Its origin is consid- through the phenylpropanoid pathway, and the concentra- ered to be the area around the source of the Irrawaddy River, tion varies among varieties or cultivars. Phenylalanine ammonia-lyase (PAL) activity is strongly related to catechin Communicated by T. Kondo production (Iwasa 1977). In an earlier study, we employed Received August 14, 2003. Accepted March 8, 2004. an RFLP analysis that used PAL cDNA as a DNA probe to *Corresponding author (e-mail: [email protected]) determine that tea PAL exists as a single gene and that there 232 Matsumoto, Kiriiwa and Yamaguchi are polymorphic differences between Japanese and Chinese teas (Matsumoto et al. 1994). Furthermore, Japanese teas show 3 multi-fragment alleles, and the frequencies of these fragments are homogenous among different regions in Japan. The PAL DNA marker is a powerful tool for the classification and evaluation of tea genetic resources (Matsumoto et al. 2002). Here, we used PAL RFLP analysis to examine plants in the remains of tea gardens around old Korean temples and at a tea farm to determine the genetic background of Korean teas and their relationship to Japanese teas.

Materials and Methods

Plant materials Sampling sites and sample numbers are shown in Fig- ure 1. Seed capsules or seeds were collected from the pre- cincts of the Korean temples of Hyang lim, Sun am, Song kwang, Zeung sim, Man yeun and Dae won, and from a tea farm at Hyoi chun myen. Seeds were sown in pots, and the germinated plants were transported to the field of the

National Institute of Vegetable and Tea Science, Kanaya, Fig. 1. Locations of the Korean sampling sites, and numbers of sam- Shizuoka, Japan, in 1994. Fresh leaves were obtained from ples from each site used for analysis. 4- or 5-year-old seedlings for DNA extraction. At the time of extraction, samples were obtained where possible from (Amersham, Hybond N+), and cross-linked to the membrane the original mother plant in the temple. The final number of with UV light. Full-length cDNA including the flanking samples was 297. Although we did not exclude the possibil- region of tea PAL (accession number D26596) (DNA ity of including samples from the same mother plant, the ra- fragment total 2.3 kb), was labelled with α-[32P]dCTP or tio of contamination was very low and we considered that it fluorescein-11dUTP by the random primer method to act as would not seriously influence our results. Genomic DNA a probe. The labelled DNA probe was denatured at 95°C for was extracted by a modification of the method of Guillermaut 2 min and then added to the membrane in hybridization buffer and Maréchal-Drouard (1992). Two grams of fresh leaves (6 × SSC, 50 × Denhardt’s solution, 0.1 mg/mL denatured was ground in a mortar and pestle with 17 mL DNA extrac- DNA, 0.1 % SDS for α-[32P]dCTP-labelled probe; 5 × SSC, tion buffer (100 mM sodium acetate [pH 4.8], 50 mM EDTA, 1/20 dilution Liquid Block, 0.1 % SDS and 5 % dextran sul- 500 mM NaCl, 2 % soluble PVP [MW 40,000], 1.4 % SDS), fate for fluorescein-labelled probe). Hybridization was per- and kept warm for 30 min at 65°C. Samples were centri- formed overnight at 65°C. After hybridization, the nylon fuged at 10,000 rpm at room temperature for 10 min, and membrane was washed at room temperature for 5 min with 2 15 mL of supernatant was transferred to a new tube. After × SSC and then at 65°C with 0.2 × SSC solution for 15 min; addition of 6.8 mL of 4 M potassium acetate (pH 4.8), the this was repeated twice. DNA fragments hybridized with the sample was put on ice for 30 min. The sample was then cen- labelled probe were detected on X-ray film. The steps for de- trifuged at 12,000 rpm at 4°C for 20 min. After the super- tection of fluorescence were performed according to the natant had been transferred to a new tube, an equal volume of manufacturer’s instructions (Amersham Pharmacia, Gene isopropanol was added to it, and the mixture of DNA fibers Image). For fragments labeled with α-[32P], films were ex- became visible upon gentle swirling. The DNA was rinsed posed under intensifying screens at −80°C for 3 to 7 days be- twice with 70 % ethanol and dissolved in 5 mL TE (10 mM fore being developed. In each experiment, the Japanese cul- Tris·HCl [pH 8.0], 1 mM EDTA). RNA in the sample was di- tivars Yabukita and Ujihikari were used for comparison. gested by addition of 100 µL RNase (10 mg/mL) and warmed at 55°C for 15 min. DNA was precipitated by addi- Results tion of 5 mL isopropanol, followed by centrifuging and a further washing in 70 % ethanol, and then finally dissolved in 1 Diversity of Korean teas growing around old temples mL TE. When the DNAs of the Japanese teas were digested by HindIII and hybridized with the PAL probe, 4 intense frag- Southern blot analysis ments, 7.3 kb (A), 6.7 kb (B), 5.9 kb (C) and 5.3 kb (D), were Fifteen micrograms of DNA was separately digested detected, for example, A and C in ‘Ujihikari’ and B, C and D using 2 restriction enzymes, HindIII and EcoRV. Digested in ‘Yabukita’ (Fig. 2). The A, B and D fragments are inherit- DNA was separated by electrophoresis in 0.8 % (w/v) agar- ed as multiple alleles at a single locus, and it is possible to ose, transferred to a positively charged nylon membrane classify Japanese tea cultivars into 6 PAL genotypes: AA, The Korean tea plants 233

AB, AD, BB, BD and DD by this marker (Matsumoto et al. and we termed it K6 as another multi-fragment allele. K6 1994, 2002). The RFLP patterns of tea plants from the Sun was detected in lanes 1, 2, 4, 5, 7, 11, 12 and 13 in Figure 3, am and Zeung sim temples are shown in Figure 2. In the and the fragments detected in the same position as B in lanes RFLP pattern of the Zeung sim teas, fragments at 2.3 kb 6, 8, 9 and 10 were derived from K1 or K2. We considered (lanes 8, 10 and 12), 2.5 kb (lane 10), 3.6 kb (lane 3, 5, 8 and the clones in lanes 7, 12 and 13 to be combinations of K6 9) and 4.1 kb (lanes 1, 2, 4, 6, 7 and 11) were strongly detect- with K1 or K2, because their fragments were more intense ed. Fragments at 3.6 kb (lanes 2, 3, 5, 12 and 13), 4.1 kb than the others. (lanes 1–11, 13) and about 9.5 kb (lanes 7 and 12) were de- The b1 fragment was detected only in lanes 9, 10 and tected in samples from the Sun am temple. In the lanes of tea 11, by EcoRV and the 3.6 kb fragment (K3) was also detect- plants possessing the 2.3 and 2.5 kb fragments, a fragment ed in the same lanes by HindIII (Fig. 3). The K3 and B alleles positioned at B (or a little lower than B) was also detected. probably have the same EcoRV recognition site; when K3 or Similarly, a fragment at C, or close to it, was also detected in the B allele is digested by EcoRV, the b1 fragment is detect- the lanes in which we found 3.6 kb and 9.5 kb fragments. ed. We found the same relationship between the 2.3 and 2.5 Since the PAL gene exists at a single locus and these frag- kb fragments after HindIII digestion and the b2 fragment af- ments were detected coincidentally, we speculate that these ter EcoRV digestion. From these fragment patterns in the were multi-fragment alleles that were not detected in the PAL RFLP analysis of Korean tea plants, we assumed the analysis of Japanese teas. Therefore, the fragments at 2.3, existence of 7 new multi-fragment alleles, shown in Table 1 2.5, 3.6 kb and about 9.5 kb were designated K1, K2, K3 and as K1 to K7. In the Korean tea plants, no fewer than 3 Japa- K7, respectively. nese types of multi-fragment allele–A, B and D–are includ- A 4.1 kb fragment was detected among the local varie- ed, so a total of at least 10 alleles exist. ties in Japan, but the allelic frequency value was quite low at 0.03 in our previous studies. We consider that the 4.1 kb Tea plants at the farm at Hyoi chun myen fragment was probably a variation at the side of the 5′ do- The RFLP patterns from HindIII digestion of the Hyoi main, as was fragment C (Matsumoto et al. 2002). The de- chun myen farm plants are shown in Figure 4. Although tectabilities of fragment C and the 3 other fragments (A, B there were only 12 samples, the allelic fragments consisted and D) detected in Japanese tea can differ with changes in of A, B and D; all samples were of the Japanese type; 8 were the PAL probe, if only the domain beside the 5′ on the PAL of PAL genotype AA (lanes 2, 3, 6, 8–11 and 12), 2 of geno- cDNA is used as a probe, only the C fragment is detected; in type AB (lanes 4 and 7), and 2 of genotype AD (lanes 1 and contrast, the 3′ domain hybridizes only to the A, B and D 5). The same results were detected on analysis of the EcoRV fragments (Matsumoto et al. 1994). Fragment C is common, digestion (data was not shown). The RFLP patterns were and it is thought that the PAL alleles detected in Japanese tea clearly different from those of the plants growing around the are composed of C–A, C–B and C–D. In the analysis of temples. Korean teas, the frequency of occurrence of the 4.1 kb fragment was higher than in the analysis of Japanese teas. Teas Comparison of gene frequencies among sampling areas that possessed a 4.1 kb fragment without a C fragment were The proportion of each PAL fragment in samples from found (Fig. 2, lanes 2, 4, 6 and 11 from the Zeung sim tem- each site is shown in Table 2. Samples from the Sun am temple, and lanes 6, 8, 9, 10 and 11 from the Sum am temple). ple (3 % total Japanese genotype) and the Hyang lim temple We also speculated that there were 2 types of PAL, one com- (5 % total Japanese genotype) consisted of PAL alleles that posed of the 4.1 kb and A fragments (4.1 kb–A), and the oth- had little in common with the Japanese types. The propor- er of the 4.1 kb and B fragments (4.1 kb–B); these were des- tions of the Japanese PAL genotype in plants from the Zeung ignated K4 and K5, respectively. With the exception of lane sim, Song kwang, Dae won and Man yeun temples were 14 12, all teas from the Sun am temple in Figure 2 possessed the %, 20.7 %, 28.6 % and 29.2 %, respectively. In contrast, 24 K5 allele, and the teas of lanes 2, 4, 6 and 11 from the Zeung alleles of 12 individual plants from the farm at Hyoi chun sim temple possessed K4 and K5 heterozygously. myen were entirely of the Japanese type, revealing that these The RFLP patterns of the tea samples from the Song plants and the temple plants had different origins. Although kwang temple are shown in Figure 3. When ‘Yabukita’ (BD the total proportion of the A fragment was 66 % in Japanese genotype) was digested by EcoRV, fragments near 3.2 kb teas examined previously (Matsumoto et al. 2002), it was (b1), 2.8 kb and 1.8 kb were visible as intense bands. The b1 only 9.9 % in the Korean temple teas, in which the K6 frag- fragment in the EcoRV digestion corresponded to the B ment, not detected in the Japanese teas, occurred at the high- fragment in the HindIII digestion, and if it has a B fragment, est mean frequency (19.9 %). the b1 fragment will surely be detected. On the other hand, for the tea clone in lane 5 in Figure 3, only 2 fragments were Discussion detected in the same positions as B and C; although this sample was similar to the BB type after HindIII digestion ac- In RFLP analysis using PAL cDNA as a probe, teas col- cording to PAL genetic typing, it did not possess the b1 frag- lected from the grounds of Korean temples showed highly ment. This fragment without b1 therefore differs from B, polymorphic patterns. Four fragments, at 2.3, 2.5, 3.6 and 234 Matsumoto, Kiriiwa and Yamaguchi

Fig. 2. PAL RFLP patterns after digestion by HindIII of Korean tea plants collected from the Sun am temple (top) and the Zeung sim temple (bottom). Lanes 1 to 13 (top) and 1 to 12 (bottom) are Korean teas, and YK (‘Yabukita’) and UH (‘Ujihikari’) are Japanese green tea cultivars with BD and AA PAL genotypes, respectively.

9.5 kb, were detected strongly among Korean teas. Because Taiwanese and Japanese; the low diversity of the Japanese tea PAL exists at a single locus, we consider these to be tea plants might be the result of intense selection and cross- multi-fragment alleles (K1, K2, K3 and K7). The 4.1 kb frag- ing from a genetically limited stock (Kaundun et al. 2000). ment was detected among local Japanese tea varieties at a Moreover, Euclidian distance values calculated from RAPD lower frequency than in Korean teas, and we concluded that analysis of 23 Korean and 25 Japanese teas have been com- there were 2 types of PAL allele, 4.1 kb–A (K4) and 4.1 kb– pared, and the genetic diversity of the Korean teas was again B (K5). The PALs of Japanese local varieties were com- higher than that of the Japanese (Lee et al. 1995). posed of 3 multi-fragment alleles, A, B and D. One of the In a morphological analysis of Korean teas used in this Korean tea fragments detected in the same position as the study, although the leaves tended to be longer and more Japanese B fragment in analysis using HindIII was shown by slender than Japanese leaves, their leaf size values fell with- EcoRV to differ from the Japanese B pattern, and we classi- in the range of variation accepted for Japanese teas (Ikeda fied it as another allelic fragment, K6. Finally, we assumed and Park 2002). Because of these very close similarities in that the Korean teas had 10 multi-fragment alleles, including leaf form, we considered that judgment by morphological 3 Japanese types, and that the genetic diversity of Korean characters alone is difficult. Relative pistil height is used as teas was therefore much greater than that of Japanese. Other an index to distinguish Japanese and Chinese teas. There are studies of the diversity of Korean and Japanese teas have re- 3 types, the L type, in which the pistil is longer than the sta- ported similar results. One RAPD analysis of green tea men; the S type, in which it is shorter than the stamen; and plants, in which 8 Japanese cultivars, 5 Taiwanese cultivars the M type, in which it is the same length as the stamen. Va- and 14 Korean lines were studied, showed clearly that diver- riety assamica and most of the Chinese native species are of sity was greatest within the Korean group, followed by the the L or M type, and almost all Japanese teas are of the M or The Korean tea plants 235

Fig. 3. PAL RFLP patterns after digestion by HindIII (top) and EcoRV (bottom) of Korean tea plants collected from the Song kwang temple. Lanes 1 to 13 are collected samples, and YK and UH are the Japanese teas ‘Yabukita’ and ‘Ujihikari’, respectively. The B fragment of YK HindIII digestion corresponds to the b1 fragment of EcoRV digestion. The 3.6-kb fragment in the Korean tea plants corresponds to b1. A similar relationship was detected between the 2.5 and 2.3 kb fragments and the b2 fragment.

Table 1. Comparison of estimated PAL fragment alleles of Korean and Japanese tea plants, based on RFLP analysis Restriction of K1 K2 K3 K4 K5 K61) K7 A2) BD fragment length 9.5 kb −−−−−−+−−− 7.3 kb (A) −−−+−−−+−− 6.7 kb (B) ++−−++−−+− 5.9 kb (C) −−+−−+++++ 5.3 kb (D) −−−−−−−−−+ 4.1 kb −−−++−−−−− 3.6 kb −−+−−−−−−− 2.5 kb −+−−−−−−−− 2.3 kb +−−−−−−−−− 1) Although Korean type K6 seemed to be the same as Japanese type B, the pattern was different from Japanese type B on EcoRV analysis. 2) Three types of fragment–A, B and D–were detected in Japanese teas. Japanese teas can therefore be divided into 6 PAL genotypes: AA, AB, AD, BB, BD and DD (Matsumoto et al. 1994).

S type (Takeda and Toyao 1980). In teas collected in Korean index is limited to flowering time or plants old enough to temples, unlike the Japanese, there are many L types, and the bloom. Therefore, development of a marker that can help teas are considered closer to the Chinese types (Ikeda and classify tea genetic resources even if the sample is too young Park 2002). PAL multi-fragment alleles of the teas were also to flower is important. different from the Japanese; in particular, the Sum am tem- AFLP (Paul et al. 1997), RAPD (Lee et al. 1995, Tanaka ple and the Hyang lim temple teas contained few Japanese and Yamaguchi 1996, Wachira et al. 1997, Kaundun et al. alleles. Although pistil height is stable and not influenced by 2000, Tanaka 2001) and RFLP (Matsumoto et al. 1994) the environment, any evaluation that uses pistil height as an markers for evaluation of tea genetic resources, RAPD for 236 Matsumoto, Kiriiwa and Yamaguchi

Fig. 4. PAL RFLP patterns after HindIII digestion of 12 samples collected from a tea farm at Hyoi chun myen, Korea, and 2 Japanese tea cultivars. Lanes 1 to 12: samples from the tea farm. YK: ‘Yabukita’, UH: ‘Ujihikari’.

Table 2. Proportions of PAL fragment alleles in Korean tea plants

No. of samples Fragment allele Collection site Unknown analyzed K1 K2 K3 K4 K5 K6 K7 A B D Sun am temple 50 0.0 1.0 13.0 0.0 41.0 26.0 4.0 3.0 0.0 0.0 12.0 Hyang lim temple 50 0.0 29.0 26.0 3.0 11.0 24.0 2.0 5.0 0.0 0.0 0.0 Zeung sim temple 47 3.2 20.2 11.7 17.0 18.1 13.8 0.0 4.3 6.4 3.2 2.1 Song kwang temple 44 3.4 24.1 12.6 1.1 10.3 23.0 4.6 13.8 0.0 6.9 0.0 Dae won temple 46 4.4 14.3 8.8 18.7 14.3 5.5 4.4 15.4 0.0 13.2 1.1 Man yeun temple 48 2.1 18.8 1.0 10.4 8.3 26.0 4.2 18.8 7.3 3.1 0.0 Total 285 2.1 17.8 12.3 8.3 17.4 19.9 3.2 9.9 2.3 4.2 3.0 Hyoi chun myen farm1) 12 0.0 0.0 0.0 0.0 0.0 0.0 0.0 83.3 8.3 8.3 0.0 1) Because the samples from Hyoi chun myen farm were different from those taken from the 6 temples, the farm samples were excluded from the calculation of the fragment allele proportions. paternity testing, and SSR (Kaundun and Matsumoto 2002) in Korea (Kim 2000). Between 1927 and 1929, Japanese tea and CAPS markers (Kaundun and Matsumoto 2003) for cul- seeds from Kumamoto, Kyoto and Ehime prefectures were tivar identification are all used in tea genetic studies, making carried to Korea to develop tea farms (Kimura 1996). Thus, DNA analysis of tea a powerful evaluation technique. In our there is historical evidence of the introduction of large quan- previous studies we indicated that PAL was a useful DNA tities of tea seed into Korea from Japan. Further, RAPD marker for evaluation of tea genetic resources and identify- analysis of 23 local Korean varieties revealed that 3 being tea cultivars (Matsumoto et al. 1994, Matsumoto et al. longed to the Japanese group, and these had presumably 2002, Kaundun and Matsumoto 2003). Also, in the current been introduced from Japan or were hybrids with Japanese study we expressed genetic diversity by the number of PAL cultivars (Lee et al. 1995). Therefore, tea originating from alleles and found higher diversity in the Korean tea popula- Japan survived and grew on tea farms in Korea. Other local tion. Our results were consistent with those of RAPD analy- varieties were probably introduced from China centuries ago sis (Lee et al. 1995, Kaundun et al. 2000) and morphological and remained around the old temples. Therefore, in Korea analysis (Ikeda and Park 2002). PAL genetic diversity there- there are 2 tea groups with different origins. fore appears to be closely related to differentiation of the tea Because Japanese and Korean teas can be crossed with plant. each other, it is possible that hybrids between the 2 groups All samples collected from the farm at Hyoi chun myen will be raised and will show high genetic diversity. In Korea, had fragments with Japanese-type patterns. Moreover, the if a new cultivar that is propagated vegetatively is developed pistil heights of sample plants were low, unlike those of the and becomes popular, then tea seedling farms will become temple plants, adding further evidence that the plants were less common and there may be a loss of genetic resources. of the Japanese type (Ikeda and Park 2002). From these Therefore, we need to further evaluate the genetic character- DNA and morphological analyses, we can say for certain istics of Korean tea plants and preserve them as soon as pos- that the teas at the Hyoi chun myen farm were derived from sible: for example, the index of cold hardiness of Korean Japanese types. Between 1910 and 1945, Japanese immi- teas is the highest among var. sinensis, including Chinese grants developed tea farms at Guanzhou, Imura and Posong and Japanese tea resources (Takeda 2002). It is easy to see The Korean tea plants 237 how teas with cold hardiness were selected through adapta- Camellia sinensis. Genome 45: 1041-1048. tion to the severe Korean winters. The breeding of Japanese Kaundun, S.S. and S. Matsumoto (2003) Development of CAPS mark- green teas has produced early budding cultivars with unique ers based on three key genes of the phenylpropanoid pathway Camellia sinensis aromas and flavors from a small genetic pool. Cold hardi- in Tea, (L.) O. Kuntze, and differentiation assamica sinensis ness is an important character in Japanese tea breeding for between and varieties. Theor. Appl. Genet. 106: 375-383. northern tea cultivation regions. Today, consumers are look- Kim, M. (2000) Sencha no kigen to hatten happyo ronbun-shu (Pro- ing for new tastes and aromas, for example those supplied by ceedings of a symposium on the origin and development of Chinese-type teas. The genetic resources of the conventional green tea). Kurofune-Insatsu, Shizuoka, Japan. p. 19-45 (in Japanese varieties are likely to be inadequate to meet these Japanese). needs, but the cold-hardy Korean teas are promising. Al- Kimura, K. (1993) Research on tea in Korea. Bull. Univ. Setouchi though ‘Okumidori’ and ‘Yamatomidori’ are regarded as Junior College 23: 39-47 (in Japanese with English summary). Japanese teas, they have long pistils and possess the K3 PAL Kimura, K. (1996) Research on the present condition of tea in Korea. fragment (Matsumoto et al. 2000). Because K3 is not detect- Bull. Univ. Setouchi Junior College 26: 29-46 (in Japanese ed in local varieties in Japan, the origin of these cultivars is with English summary). either Korea or China. Some of the other characters of these Lee,S.H., H.S.Choi, R.H.Kim, H.Y.Lee and I.S.Nou (1995) Identifi- 2 cultivars are different from those of conventional Japanese cation of Korean wild tea plants and Japanese green tea cultivars using RAPD markers. J. Kor. Tea Soc. 1: 129-148. teas. The budding period of ‘Yamatomidori’ is the latest Matsumoto, S., A. Takeuchi, M. Hayatsu and S. Kondo (1994) Molecu- among Japanese green tea cultivars and local Japanese vari- lar cloning of phenylalanine ammonia-lyase cDNA and classi- eties, and ‘Okumidori’ is also a late budding cultivar whose fication of varieties of tea plants (Camellia sinensis) using tea market share is increasing because of its good quality. Even PAL cDNA probe. Theor. Appl. Genet. 87: 671-675. if overseas genetic resources need to be used as breeding Matsumoto, S., K. Kiriiwa and Y. Takeda (2000) Analysis of genetic parents, such examples show that new cultivars of good diversity among Chinese tea (Camellia sinensis) using RFLP quality can be produced. We now need to plan breeding marker and detection of difference of it from Japanese. Breed. programs that select the appropriate resources for cross- Res. 2 (Suppl. 2): 209 (in Japanese). breeding to fully utilize the genetic diversity of teas. Matsumoto, S., K. Kiriiwa and Y. Takeda (2002) Differentiation of Japanese green tea cultivars as revealed by RFLP analysis of Acknowledgment phenylalanine ammonia-lyase DNA. Theor. Appl. Genet. 104: 998-1002. Paul, S., F.N. Wachira, W. Powell and R. Waugh (1997) Diversity and We thank Dr. Park Young-Goo for his help in collect- genetic differentiation among populations of Indian and Ken- ing the plant material used in this study and for useful infor- yan tea (Camellia sinensis (L.) O. Kuntze) revealed by AFLP mation about Korean teas. markers. Theor. Appl. Genet. 94: 255-263. Takeda, Y. (2002) Studies on variation in genetic resources of tea in Literature Cited Japan and application to tea breeding. Bull. Natl. Inst. Veg. Tea Sci. Japan 1: 97-180 (in Japanese with English summary). Choi, H.K. (2000) Tea breeding and cultivation in Korea. J. Korean Takeda, Y. and T. Toyao (1980) Identification and classification of the Tea Soc. 6: 121-137. green tea varieties by the morphological characters of flower Chun, B.S. (2000) The history of tea related culture in Korea. J. Korean organs. Tea Res. J. 52: 1-6 (in Japanese with English summary). Tea Soc. 6: 29-40. Tanaka, J. (2001) Relationship between tea breeding and genetic analy- Eden, T. (1958) Tea. Longmans, London, UK. 1 p. sis by DNA markers. Proceedings of International Conference Guillermaut, P. and L. Maréchal-Drouard (1992) Isolation of plant on O-Cha Culture and Science. Shizuoka, Japan, Oct. 5–8, DNA: a fast, inexpensive, and reliable method. Plant Mol. Biol. 2001. The Organizing Committee of the 2001 International Rep. 10: 60-65. Conference on O-CHA (tea) Culture and Science. Session II Ikeda, N. and Y.G. Park (2002) Morphological characteristics of Korean Production, p. 13-14. wild tea populations. Breed. Res. 4: 193-200 (in Japanese with Tanaka, J. and S. Yamaguchi (1996) Use of RAPD markers for the English summary). identification of parentage of tea cultivars. Bull. Natl. Res. Inst. Iwasa, K. (1977) Biosynthesis of catechins in tea plant. Bull. Natl. Res. Veg. Ornam. Plants and Tea Japan 9: 31-36 (in Japanese with Inst. Tea. 13: 101-126. English summary). Kaundun, S.S., A. Zhyvoloup and Y.G. Park (2000) Evaluation of the ge- Wachira, F.N., W. Powell and R. Waugh (1997) An assessment of genetic diversity among elite tea (Camellia sinensis var. sinensis) netic diversity among Camellia sinensis L. (cultivated tea) and genotypes using RAPD markers. Euphytica 115: 7-16. its wild relatives based on randomly amplified polymorphic Kaundun, S.S. and S. Matsumoto (2002) Heterologous nuclear and DNA and organelle specific STS. Heredity 78: 603-611. chloroplast microsatellite amplification and variation in tea,