2021 年 90 巻 2 号 p. 223-231
Evergreen azalea is one of the most important ornamental shrubs and pot plants in temperate zones worldwide. In Japan, hundreds of azalea cultivars have been bred based on the genetic diversity of wild species and various accumulated mutants since the middle of the 17th century. Japanese cultivar groups such as Edo-kirishima, Kurume-tsutsuji, Ryūkyū-tsutsuji, Hirado-tsutsuji, and Satsuki have been developed by selection and crossing, and many cultivars have been exported to Western countries and utilized as breeding materials for pot and garden azalea. Rhododendron ripense Makino, which grows on riverside rocks and is endemic to Japan, is one of the best ornamental species because of its high adaptability to environmental conditions. We have focused on the genetic contribution of this wild species to evergreen azalea cultivars, and developed a PCR-RFLP identification marker of R. ripense cpDNA based on a species-specific sequence of the trn L-F region. The R. ripense cpDNA specific marker has been in Japanese large-flowered groups, all Ryūkyū and Ōkirishima cultivars, and half of all Hirado cultivars have the R. ripense cpDNA type. Most Japanese small flower cultivars, such as Edo-kirishima, Kurume and Satsuki have non-R. ripense type cpDNA. Italian large-flowered cultivars also tend to be the R. ripense cpDNA type. Furthermore, all pot azalea cultivars of the Indian and Simsii groups possess R. ripense type cpDNA. These results clarified the cytoplasmic contribution of R. ripense not only to Japanese large flower cultivars, but also to Western azalea cultivars. Although R. simsii has been considered to be the main ancestral species of pot azalea, R. ripense should be recognized as the cytoplasmic parent of these cultivars. The ornamental value and adaptive environmental trait originating from R. ripense should be reviewed to elucidate the development history of evergreen azalea cultivars.
In the genus Rhododendron L. (Ericaceae), which comprises over 1000 species mainly in the Northern hemisphere, about 50 Rhododendron species are native to Japan (Kurashige, 2017). The endogenous species in the subgenus Tsutsusi section Tsutsusi are indispensable genetic resources for evergreen azalea (= “tsutsuji” in Japanese) cultivars used as ornamental shrubs or pot plants in temperate regions around the world. Several endemic Japanese species such as R. kaempferi Planch., R. kiusianum Makino, R. indicum (L.) Sweet, R. eriocurpum (Hayata) Nakai, R. macrosepalum Maxim., R. ripense Makino, and R. scabrum G. Don have high ornamental value and hundreds of azalea cultivars have been selected from natural populations and bred in gardens intensively since the Edo era (1603–1867) (Ito and Creech, 1984; Kobayashi, 2013, 2016; Kurashige and Kobayashi, 2008).
The small-flowered cultivar groups such as Edo-kirishima and Kurume-tsutsuji were developed based on R. kaempferi, R. kiusianum, etc. and the Satsuki cultivar group was developed based on R. indicum and R. eriocarpum. Meanwhile, the large-flowered cultivar groups such as Ryūkyū-tsutsuji, Ōkirishima and Hirado-tsutsuji were developed based on R. macrosepalum, R. ripense, R. scabrum etc. (Kobayashi, 2013, 2016; Tamura, 1963). From the 17th to 19th century, many of these varieties were exported to Western countries as new Oriental ornamentals. Rhododendron simsii Planch., native to China and described as Azalea indica, was also introduced to England (1808) and the early breeding of azaleas in Europe began. All these early azaleas and their hybrids were known as Indian Azaleas and they were popular greenhouse and indoor pot plants. After that, Belgian Indian hybrids were developed mainly in Belgium (Galle, 1987; Heursel, 1996). The genetic participation of wild azalea in the development of evergreen azalea cultivars has been investigated using DNA markers (De Riek et al., 2008; Kobayashi et al., 2000, 2003, 2007, 2008, 2017; Miyano et al., 2013; Scariot et al., 2007; Yamamoto et al., 2019). However, the detailed ancestral relationship between Japanese wild species and traditional cultivars still has some unsettled issues.
Rhododendron ripense Makino (= “Kishi-tsutsuji” in Japanese), which belongs to the section Tsutsusi subsection Scabra, is distributed in rocky riverside habitats of Western Japan, mainly in the San-in and Shikoku regions (Kobayashi et al., 2008; Kondo et al., 2009). This endemic Japanese species is the important ancestral parent of evergreen azalea cultivars, in particular Ryūkyū-tsutsuji [R. × mucronatum (Blume) G. Don] (Chamberlain and Rae, 1990; Inobe, 1971). These large-flowered azaleas are the best ornamental species because of their high adaptability to environmental conditions (Kobayashi et al., 2010a, b; Scariot et al., 2013).
Markers developed from chloroplast DNA (cpDNA) are desirable for taxonomic and phylogenetic relationship studies at various systematic levels (Chung and Staub, 2003) as the cpDNA evolution rate is slower than nuclear DNA (Wolfe et al., 1987). The PCR-RFLP method is very effective and convenient for determining molecular phylogeny (Yonemori et al., 1996) and when we survey the relationships within closely related species, we can observe the spacer regions between coding regions of cpDNA (Tsumura et al., 1995). In a previous study of introgressive hybridization between R. kiusianum and R. kaempferi on the mountains in Kyusyu, PCR-RFLP analysis of cpDNA could detect specific bands for the two species in the 16S rDNA region. Populations of interspecific hybrids that have varietal flower colors were composed of individuals that had cpDNA of either R. kiusianum or R. kaempferi. Past cytoplasmic introgression from R. kaempferi to R. kiusianum is suggested because the morphologically identified R. kiusianum individuals possessed cpDNA of R. kaempferi (Kobayashi et al., 2000, 2007). This cpDNA PCR-RFLP marker was also useful for determining the cytoplasmic origin of Edo-kirishima, Kurume and Miyama-kirishima cultivars (Kobayashi et al., 2003).
In this study, we focused on the genetic relationship of R. ripense with Japanese and Western azalea cultivars. We developed an R. ripense cpDNA specific PCR-RFLP marker based on a restriction site mutation of the trn L-F region. Using this cpDNA marker, we could evaluate the cytoplasmic genetic contribution of R. ripense to evergreen azalea cultivars. The contribution of this wild azalea species to the development of evergreen azalea cultivars was also considered by examining the long history of azalea cultivars.
Evergreen azalea species and most cultivars used in this study were collected from wild habitats in Japan, Niigata Prefectural Botanical Garden, Koishikawa Botanical Garden and the azalea resources collection of the Plant Breeding Laboratory of the Faculty of Life and Environmental Sciences of Shimane University (Tables 1, 2, and 3). Total genomic DNA was extracted from −80°C frozen fresh leaves of each plant by a modified CTAB method (Kobayashi et al., 1998). Some DNA samples of Western azalea cultivars were obtained from the University of Turin, Italy and ILVO, Belgium.
Evergreen azalea species of the section Tsutsusi used in this study and their PCR-RFLP patterns of cpDNA.
Japanese azalea cultivars used in this study and their PCR-RFLP patterns of cpDNA.
Western azalea cultivars used in this study and their PCR-RFLP patterns of cpDNA.
To develop a species-specific chloroplast DNA marker, sequences of several regions such as mat K, trn L-trn F, trn S-trn G, and atp B-rbc L, which were used for genus Rhododendron classification (Kurashige et al., 1998), were researched in the NCBI database <https://www.ncbi.nlm.nih.gov/>. The sequences of trn L-trn F (AB105247) and trn S-trn G (AB105291) of R. ripense have been registered in the database. We compared and searched for differences between wild species that were related to the establishment of evergreen azalea cultivars. One base mutation which was conformed at the digestion site of endonuclease Aci I was discovered in the trn L-trn F region in R. ripense and other species (Fig. 1). In order to detect the R. ripense cpDNA by PCR-RFLP analysis, a PCR primer set was designed to amplify the trn L-trn F region. The reliability of the R. ripense cpDNA specific PCR-RFLP marker was confirmed by using 222 individuals from 16 species belonging to the subgenus Tsutsusi (Table 1).
The investigated trn L-trn F region by PCR-RFLP analysis and the primer set used to amplify the region.
All DNA samples of 165 Japanese cultivars (Table 2) and 96 Western cultivars (Table 3) were analyzed using the newly developed PCR-RFLP marker. PCR was performed in 20 μL reaction mixtures containing 10 ng gDNA, 1 × reaction buffer, 0.2 mM dNTPs, 1.25 U of Blend Taq (Toyobo Co., Ltd., Osaka, Japan), and 0.2 μM of each primer. The reaction conditions for PCR amplification were as follows: preheating at 94°C for 3 min; 35 cycles of denaturation at 94°C for 30 s, annealing at 50°C for 30 s, and extension at 72°C for 30 s; and final extension at 72°C for 10 min. using a thermal cycler (PC-320; ASTEC Co., Ltd., Fukuoka, Japan). PCR products were digested with endonuclease Aci I (New England Biolabs Japan Inc., Tokyo, Japan) at 37°C for 2 hrs. The restriction fragments were separated by electrophoresis in 2% agarose gel (Nippon Genetics Co., Ltd., Tokyo, Japan) and 0.5 × TBE buffer at 100 V for 40 min and photographed under UV light after staining with ethidium bromide.
Using the developed PCR primer set, 340 bp of the trn L-trn F region could be amplified in all tested wild azalea samples. All PCR products of 152 individuals from 14 species except for R. ripense and a part of R. transiens Nakai were digested to 300 and 40 bp fragments by Aci I. On the other hand, no PCR products from 147 individuals of R. ripense from 14 river habitats were digested by Aci I. In addition, five of 19 R. transiens individuals that have a genetic relationship with R. ripense (Miyano et al., 2013) were not digested by Aci I either (Table 1). Therefore, this PCR-RFLP marker could distinguish R. ripense from other evergreen azalea species. In the following study, DNA samples of several cultivar groups were investigated using this R. ripense cpDNA specific PCR-RFLP marker (Fig. 2).
PCR-RFLP band patterns of the trn L-F region digested by Aci I in wild species. Lane M, 100 bp DNA ladder; Lane 1, R. ripense; Lane 2, R. scabrum; Lane 3, R. yakuinsulare; Lane 4, R. macrosepalum; Lane 5, R. yedoense; Lane 6, R. eriocarpum; Lane 7, R. indicum; Lane 8, R. serpyllifolium; Lane 9, R. kaempferi; Lane 10, R. tosaense; Lane 11, R. kiusianum; Lane 12, R. simsii; Lane 13, R. nakaharae; Lane 14, R. oldhamii.
The PCR-RFLP band patterns of several cultivars belonging to each group are shown in Figure 3 and the results for all cultivars are listed in Tables 2 and 3.
PCR-RFLP band patterns of the trn L-F region digested by Aci I in cultivars. Lane M, 100 bp DNA ladder; Lane 1, Edo-kirishima ‘Honkirishima’; Lanes 2–3, Kurume-tsutsuji ‘Tagonoura’ and ‘Hanaasobi’; Lane 4, Yama-tsutsuji ‘Kuchibeni’; Lane 5, Miyama-kirishima ‘Harunoumi’; Lane 6, Hannō-tsutsuji ‘Amagibenichōju’; Lanes 7–8, Satsuki ‘Ōryū’ and ‘Senbazuru’; Lane 9, Mochi-tsutsuji ‘Shideguruma’; Lane 10, Chōsen-yamatsutsuji ‘Yodogawa’; Lane 11, Kishi-tsutsuji ‘Momoka’; Lane 12, Ryūkyū-tsutsuji ‘Sekidera’; Lane 13, Ōkirishima ‘Ōmurasaki’; Lanes 14–15, Ōyama-tsutsuji ‘Asukagawa’ and ‘Obikinoyūbae’; Lanes 16–17, Hirado-tsutsuji ‘Rashōmon’ and ‘Seibo’.
In the small-flowered cultivar groups from Japan, all cultivars of Edo-kirishima and Kurume-tsutsuji, except for ‘Tago-no-ura’, Yama-tsutsuji (R. kaempferi), and Miyama-kirisima (R. kiusianum), had non-R. ripense type cpDNA. All Satsuki cultivars except ‘Ōryū’ and most of the Hannō-tsutsuji (R. × hannoense) cultivars were also the non-R. ripense type. In the large-flowered cultivar groups, all cultivars of Kishi-tsutsuji (R. ripense), Ryūkyū-tsutsuji, and Ōkirishima, were of the cpDNA of R. ripense type. In the Hirado-tsutsuji and Ōyama-tsutsuji groups (R. transiens), cultivars had the cpDNA type of both groups (Table 2). In Western azalea cultivars, all pot azalea cultivars in the Indian and Simsii groups possessed R. ripense cpDNA. Some Italian Japonica group cultivars were also the R. ripense type. Most Kurume and obtusum cultivars, and the Italian groups were the non-R. ripense type (Table 3).
The PCR-RFLP marker developed in this study could distinguish R. ripense from other species in the subgenus Tsutsusi section Tsutsusi that is related to cultivar development of evergreen azalea. Rhododendron macrosepalum, the most closely related species to R. ripense (Yamazaki, 1996) could also be identified by this digestion site mutation in the trn L-trn F region. Using this R. ripense cpDNA specific PCR-RFLP marker, we could investigate the cytoplasmic genetic contribution of R. ripense to the development of evergreen azalea cultivars.
All R. ripense and Ryūkyū-tsutsuji (R. × mucronatum) cultivars had 340 bp trn L-trn F fragments without a digestion site, i.e. R. ripense type cpDNA. Meanwhile, all R. macrosepalum cultivars (Yamazaki and Yamazaki, 1969) digested 300 and 40 bp fragments, i.e. of non-R. ripense type cpDNA. Ryūkyū azalea are related to R. ripense and R. macrosepalum, and are likely a hybrid of both species (Inobe, 1971; Yamazaki and Yamazaki, 1972). Chamberlain and Rae (1990) mentioned R. mucronatum (Blume) G. Don var. mucronatum as the widely cultivated white form of R. mucronatum var. ripense (Syn. R. ripense) and this may occur in the wild as the albino form of R. ripense. This white form of R. ripense corresponds to ‘Shiro-ryūkyū’. The results of this study indicated that the cytoplasmic parent of Ryūkyū-tsutsuji is R. ripense and hence supports the study by Chamberlain and Rae (1990).
‘Ōmurasaki’ (R. × pulchrum) and its sport cultivars which are classified as the Ōkirishima group also had cpDNA of the R. ripense type. These cultivars have large and brightly colored flowers and are often confused with the Hirado-tsutsuji cultivar group. However, the environmental tolerance traits of ‘Ōmurasaki’ such as being winter hardy and lime soil resistant are remarkable and they are the most important garden azalea in Japan (Kobayashi, 2016). The high adaptability to different environmental conditions originate from the traits of R. ripense, which lives in a harsh natural habitat on riverside rocks (Kobayashi et al., 2008, 2010a, b).
Hirado-tsutsuji cultivars were developed based on R. scabrum which has large red flowers, and other azaleas such as R. ripense, R. × mucronatum, and R. × pulchrum contributed to the wide variety of their flower colors and morphologies (Tamura, 1963). This established process was strengthened by flower color pigment composition and the biosynthesis gene of Hirado azalea cultivars (Meanchaipiboon et al., 2020). The results for the Hirado cultivar group in which half of the cultivars had R. ripense cpDNA also support its breeding history.
Some Ōyama-tsutsuji (R. transiens) cultivars also had R. ripense cpDNA. This result matched a report of microsatellite analysis that suggested besides R. kaempferi, R. ripense or Ryūkyū-Tsutsuji (R. × mucronatum) were related to the origin of R. transiens (Miyano et al., 2013). The results of this study provide convincing evidence for the hybrid origin of these azaleas.
All Edo-kirishima, Kurume-tsutsuji, Yama-tsutsuji, Miyama-kirishima cultivars except Kurume ‘Tago-no-ura’ had cpDNA of the non-R. ripense type. All the Western Kurume-obtusum cultivars also produced the same result because these cultivars were developed using the breeding parents in these groups (Galle, 1987). In previous PCR-RFLP analysis of the 16S rDNA region of cpDNA, most Edo-kirishima and Kurume cultivars had the banding pattern of R. kaempferi and some Kurume and Miyama-kirishima cultivars possessed that of R. kiusianum (Kobayashi et al., 2003). These results support the development history of these groups, suggesting they originated from R. kaempferi and R. kiusianum in southern Kyushu, Japan (Kobayashi et al., 2000). Some Kurume cultivars including ‘Tago-no-ura’ suggested a genetic relationship with R. × mucronatum based on observations of dorsal leaf surface characteristics (Okamoto et al., 2000). The results for ‘Tago-no-ura’ with the cpDNA type of R. ripense supports its genetic relationship with Ryūkyū-tsutsuji (R. × mucronatum) or R. ripense.
Satsuki cultivars which were developed based on R. indicum and R. eriocarpum, had cpDNA of the non-R. ripense type. Some cultivars were bred by crossing with pot azalea cultivars in order to produce gorgeous flowers (Galle, 1987). ‘Ōryū’ which is of the single R. ripense type in the Satsuki group is known as a hybrid between Satsuki and pot azalea cultivars, and the results reflected its hybrid origin.
Rhododendron × hannoense Nakai is presumed to be a hybrid between R. kaempferi and R. indicum which belong to the section Tsutsusi (Yamazaki, 1996). In this study, three out of 8 Hannō-tsutsuji cultivars (R. × hannoense) had cpDNA of the R. ripense type. According to the phylogenetic tree of cluster analysis using SSR marker data, wild R. × hannoense was clustered with R. indicum, although R. × hannoense cultivars were in the subsection of Scabra species (Yamamoto et al., 2019). The genetic relationship of these cultivars with R. ripense or Ryukyu-tsutsuji (R. × mucronatum) showing they belong to the subsection Scabra was confirmed by the three cultivars having R. ripense type cpDNA.
In Italy, azaleas were introduced from other European countries to parks and historic gardens in the Lake Maggiore area, Piedmont (Noth-West Italy) during the 19th century. Over the years they were bred and selected by growers of Verbania, giving rise to a diverse range of varieties, considered locally to be divided into three groups, according to the phenotype: Indica, including plants characterized by large flowers; Amoena, azaleas with tiny purple flowers and hose in hose; Japonica, a morphologically intermediate group between the two (Scariot et al., 2007). Some Italian large-flowered cultivars in the Japonica group were of the R. ripense cpDNA type and they have been genetically affected by Indian and Simsii hybrids.
As for the Western history of evergreen azaleas, Heursel (1996) described as follows. Evergreen azaleas were introduced into Europe from the 17th century and R. indicum, R. simsii, and R. mucronatum were the first to arrive. These ‘Indian azaleas’ were grown as tender greenhouse plants and became very popular in the 1840s and 1850s. From the 1860s, the development of Belgian ‘Indians’ that were both hardy and tender pot azalea continued in Belgium, Holland and Germany. The parents of the modern large-flowered evergreen azaleas are assumed to be three wild species; R. indicum, R. scabrum, and R. simsii, and a traditional cultivated variety; R. mucronatum. R. mucronatum has played a part in the parentage of the Indica Hybrids and the development of winter-hardy azalea (Heursel, 1996). Galle (1987) also mentioned the major parents of Belgian Indians were three forms of R. simsii collected by Robert Fortune in a Shanghai nursery and sent to England in 1851 under the name of ‘Indica’. These plants were named ‘Vittata’, ‘Vittata Puncata’, and ‘Vittata Bealii’, all with white flowers striped and flecked with red. By 1854, these three varieties reached Belgium, launching an enormous breeding and growing program (Galle, 1987).
All pot azalea cultivars of Indian (Indica) and Simsii hybrids have cpDNA of R. ripense. The azalea cultivars used in this study included old and important cultivars in azalea development history. The results of this study clarified the cytoplasmic parent of Western pot azalea is R. ripense. Scariot et al. (2007) reported using AFLP analysis that R. simsii is more distantly related to the Belgian azaleas in agreement with previous research on AFLP and matK sequences (Dendauw et al., 2002). R. simsii is generally considered to be the most important ancestor of Belgian azaleas; however, the results of these studies indicated that R. ripense is one of the most important species in the development of Belgian pot azalea.
The relationship between chloroplast and environmental tolerance has been noted in several reports. The chloroplast protein CEST induces tolerance to multiple environmental stresses and reduces damage in transgenic Arabidopsis (Yokotani et al., 2010). The chloroplast is of utmost importance for cold acclimation and acquisition of freezing tolerance (Trentmann et al., 2020). The cytoplasm genomic contribution of R. ripense, presumably through R. mucronatum for pot azalea development, is likely related with breeding objectives such as winter hardy in the adaptation to environment stress in Western countries (De Riek et al., 2018; Demasi et al., 2017).
In this study, we developed an R. ripense cpDNA specific maker which was able to identify R. ripense from other species of the section Tsutsusi. Using this PCR-RFLP marker, the possession of R. ripense cpDNA in evergreen azalea cultivars was investigated comprehensively. As a result, the contribution of the R. ripense chloroplast genome was revealed in Japanese large flower cultivars and also Western pot azalea cultivars. The ornamental value and environmental adaptation traits originating from R. ripense made a considerable contribution in the development of these cultivars. Rhododendron ripense should be re-evaluated as one of the most important ancestral species in the development history of evergreen azalea cultivars.
We are grateful to Koishikawa Botanical Gardens, University of Tokyo for providing us with some of the azalea cultivars. The authors also thank the faculty of Life and Environmental Sciences in Shimane University for financial support in publishing this report.
