ORIGINAL ARTICLES
Origins of Hosta Cultivars Based on Sequence Variations in Chloroplast DNA
2015 Volume 84 Issue 4 Pages 350-354
Details
2015 Volume 84 Issue 4 Pages 350-354
We analyzed the sequence variations in chloroplast DNA (cpDNA) for 31 Hosta cultivars, which may have been derived from Hosta sieboldiana and H. albomarginata in Japan, in order to elucidate the population origins of Hosta cultivars in their native habitats. In a previous study, we identified 39 haplotypes of cpDNA with specific regional features in the native habitats (16 and 23 in H. sieboldiana and H. albomarginata, respectively). These regional features of cpDNA variations were used to explore the maternal population origins of Hosta cultivars. We found that multiple Hosta cultivars may have originated from H. sieboldiana and H. albomarginata. We also observed some additional haplotypes that had not been detected previously. It is possible that these haplotypes originated from populations that were not investigated in previous studies, or that mutations occurred in the cultivar sequences after breeding.
Hosta is a genus of the family Asparagaceae distributed in East Asia and Russia (Takahashi, 2002; Tamura and Fujita, 2013). There are two species distributed in China (Chen and Boufford, 2000), while Chung and Chung (1988) recognized 11 species and two varieties of the genus in Korea. On the other hand, Fujita (1976) recognized 18 species and seven varieties of the genus Hosta in Japan. The Japanese Archipelago may be the center of speciation of the genus Hosta since the majority of species are distributed in Japan.
Over 100 species, varieties and cultivars of the genus Hosta have been described globally to date. It is believed that the Chinese species, H. plantaginea and H. ventricosa, were introduced into Europe approximately 230 years ago, and these were the first Hosta species cultivated in Europe (Grenfell, 1981). The Hosta cultivars are popular for landscaping and gardening in many countries. In addition, many cultivars are being traded, especially in Europe. In Japan, horticulture and garden design of Hosta cultivars flourished during the Edo period (Schmid, 1991). At this time, many Hosta cultivars originating from Japan were bred and improved (Nakamura, 2010). Although there are many Hosta cultivars on the market, the origin of these species remains unclear. Hosta sieboldiana and H. albomarginata are the two most common and widespread herbaceous species in Japan and may be the origins of some cultivars of Hosta in the Japanese Archipelago (Fujita, 1976).
The origins of cultivars of Primula sieboldii, a popular species in Japan, were examined by analyzing chloroplast DNA (cpDNA) variations, suggesting that most cultivars were derived from wild populations (Honjo et al., 2008). From the P. sieboldii cultivars, 10 cpDNA haplotypes were identified, three of which were newly detected and not present in wild populations. Honjo et al. (2008) determined that the majority of the cultivars had been improved by intraspecific crossing among P. sieboldii originating from the Asama-Arakawa populations and other cultivars from other areas of Japan. Similar to cultivars of P. sieboldii, the Hosta cultivars may have originated from different regions of Japan.
cpDNA variations have often been used to trace cultivar origins (Honjo et al., 2008; Tanaka et al., 2013). We previously analyzed sequence variations in two noncoding regions of cpDNA in over 100 wild populations of H. sieboldiana and H. albomarginata in Japan (Lee and Maki, 2013), and identified 39 haplotypes of cpDNA with specific regional features in these two species (16 and 23 in H. sieboldiana and H. albomarginata, respectively). These regional cpDNA variations can be used to explore the maternal origins of Hosta cultivars. In this study, we focused on Hosta cultivars that may be derived from the two species H. sieboldiana and H. albomarginata, which are native to Japan, and explored the population origins of these cultivars based on cpDNA haplotypes.
A total of 31 cultivars of the genus Hosta were obtained from a plant supplier, Touhou Botanical Gardens (Table 1). Although many cultivars are known for Hosta, we excluded the cultivars that are considered to have derived from species other than H. sieboldiana and H. albomarginata in this study. The horizontal to vertical ratios of the leaves in typical H. sieboldiana and H. albomarginata are 1.09–1.69 and 2.65–4.25, respectively. We tentatively categorized these cultivars into the two species, H. sieboldiana and H. albomarginata, based on the values of the horizontal to vertical ratio of the leaves and the base form (cordate or truncate) of the leaves (Fig. 1). Leaf materials were sampled from the cultivars in the spring and stored in an ultra-cold freezer (−70°C) in Ziplock plastic bags until DNA extraction.
The phenetic type, the cpDNA haplotype, and the estimated species of origin for each Hosta cultivar compared with the results of Lee and Maki (2013).
Two types of Hosta cultivar based on the morphology of leaves. (a) Hosta ‘Guacamole’: Hosta sieboldiana type. (b) Hosta ‘Shiro-Kapitan’: H. albomarginata type.
Genomic DNA was extracted from the leaves according to the CTAB method of Doyle and Doyle (1987). PCR amplifications were performed in a total reaction volume of 15 μL containing 10–20 ng of total DNA, 0.15 μM each primer, 0.1 mM deoxynucleoside triphosphates (dNTPs), 50 mM KCl, 2 mM MgCl2, 10 mM Tris-HCl (pH 8.3), and 0.375 units of Taq DNA polymerase (Amplicon Inc., Irvine, CA, USA). To examine variations in cpDNA in the Hosta cultivars, we amplified two noncoding regions (1380 bp) of cpDNA, namely, the trnS(GCU)–trnG(UCC) spacer and the trnL–rpl32F spacer, using the primers reported by Hamilton (1999) and Taberlet et al. (1991). These two non-coding regions were the same as those used in a previous study (Lee and Maki, 2013).
Double-stranded DNA was amplified by PCR with incubation at 94°C for 3 min, followed by 30 cycles of incubation at 94°C for 30 s, 55°C for 30 s (the trnL–rpl32F intergenic region) or 60°C for 60 s [the trnS(GCU)–trnG(UCC) intergenic region], and 72°C for 30 s, with final extension at 72°C for 5 min. After amplification, the PCR products were purified using Gene Clean Kit II (MP Biomedicals, Irvine, CA, USA). We sequenced the purified PCR products using a BigDyeTM Terminator Cycle Sequencing Ready Reaction Kit (Applied Biosystems, Foster City, CA, USA) and analyzed them on Genetic Analyzer models 3100 and 3700 (Applied Biosystems) according to the manufacturer’s protocol. Sequencing was conducted from both ends using the same primers as for the trnL–rpl32F spacer. We sequenced as long a region as possible from each side because we could not sequence the total range of trnL–rpl32F. We used several internal sequencing primers in trnS(GCU)–trnG(UCC) in line with previous reports (Lee and Maki, 2013).
Phylogenetic analysesAll sequences were aligned using ClustalW (Thompson et al., 1994), and indistinctly aligned regions caused by indels were corrected manually following the method of Lee and Maki (2013). The cpDNA haplotypes were determined based on these aligned sequences, and a haplotype network based on statistical parsimony was created to evaluate possible relationships among the haplotypes using TCS 1.06 (Clement et al., 2000). All indels were treated as point mutations and weighted evenly with other mutations. All haplotypes of the two Hosta species, H. sieboldiana and H. albomarginata, detected by Lee and Maki (2013), were also analyzed together with the haplotypes detected in this study.
On the basis of approximately 1380 bp of the two noncoding regions of cpDNA, seven substitutions, three indels and 11 types of polyT tracts (repeated sequences of T) were detected among the examined sequences of the Hosta cultivars (Table 2). All of the sequences obtained were deposited in DDBJ (accession nos. AB987835–AB987896). A total of 13 haplotypes were detected among these cultivars. We compared the haplotypes of the cultivars to those in natural populations of H. sieboldiana and H. albomarginata, which were detected by Lee and Maki (2013). The genetic relationships among the haplotypes of the Hosta cultivars and two natural populations of both Hosta species were revealed using the parsimony network (Fig. 2). We identified two types of origin for the Hosta cultivars; some were the same as or similar to H. sieboldiana haplotypes, and the others were the same as or similar to H. albomarginata haplotypes. Among these haplotypes, three and six were the same as H. sieboldiana and H. albomarginata, respectively, while four haplotypes were detected for the first time in this study (Table 2). Among these unique haplotypes, one and three were phylogenetically similar to H. sieboldiana and H. albomarginata, respectively.
Polymorphic sites and cpDNA haplotypes based on sequences of two noncoding regions from the Hosta cultivars.
Statistical parsimony network of the cpDNA haplotypes of the Hosta cultivars with Hosta sieboldiana and H. albomarginata. The colored haplotypes were found in both matched between the natural populations of H. sieboldiana and H. albomarginata of Lee and Maki (2013) and the cultivars examined in this study. “HG” means the unique cpDNA haplotypes of the Hosta cultivars.
Over half of the cultivars (27 out of 31) contained haplotypes of H. albomarginata (Table 1). The majority of cultivars had haplotype b, which is found in wild populations in Kinki, Chubu and Hokkaido. Four of the cultivars possessed haplotype c, which has been detected in the Tohoku region. Three other cultivars had haplotypes f, j, and q, which are distributed in the Shikoku, Chubu, and part of the Tohoku regions, respectively. Only three cultivars possessed the haplotypes of H. sieboldiana (Table 1); the samples HG2, HG4, and HG8 had haplotypes H, F, and J, respectively. Haplotype H was detected on the Pacific side of the Japanese Archipelago, and haplotype F was found only in Yamagata Prefecture in the Tohoku region. As inferred from the name Hosta ‘Sagae’ (HG4), this cultivar originated in Sagae, a city in Yamagata Prefecture. On the other hand, haplotype J was detected from the natural population in the Chubu region, Tokushima and Gifu Prefecture in Honshu.
We also found that some of the cultivars contained unique haplotypes that were not detected in the natural populations examined in the previous study. Two cultivars (HG5 and HG24) showed haplotypes closely related to haplotype b and one cultivar (HG3) had a haplotype closely related to haplotype j of H. albomarginata (Fig. 2). The haplotype of HG5 was also genetically similar to haplotype t, which was detected only in Fukui Prefecture in Honshu. However, one cultivar (HG14) showed a haplotype closely related to haplotype J of H. sieboldiana.
Comparison between phenetic types and cpDNA haplotype resultsTotals of 21 and 10 cultivars were categorized into H. sieboldiana type and H. albomarginata type, respectively, based on leaf morphology (Table 1). By contrast, according to the cpDNA haplotype results, only four of 31 cultivars possessed the haplotypes of H. sieboldiana. Only 12 cultivars coincided phenetically with the cpDNA haplotypes.
We found that the majority of Hosta cultivars had the same haplotypes as found in natural populations of H. albomarginata (haplotypes b, c, f, v, and q) and some cultivars had the haplotypes found in natural populations of H. sieboldiana (haplotypes H, F, and J). These results suggest that the majority of Hosta cultivars investigated in this study were bred from plants that originated from several natural populations of H. sieboldiana or H. albomarginata.
Some of the haplotypes were not detected in previous phylogeographic studies (Lee and Maki, 2013). It is considered that natural populations with these haplotypes were not sampled in our previous study, or that the Hosta cultivars may retain genetic diversity lost from wild populations. Another possible explanation is that spontaneous mutations occurred in the ancestral cultivar and the mutations might have accumulated during breeding in the cultivars.
The majority of Hosta cultivars contained cpDNA derived from H. albomarginata, and cultivars with cpDNA from H. sieboldiana were less common, suggesting that H. albomarginata may have been used more generally to breed cultivars than H. sieboldiana. However, several inconsistencies were observed between the morphological and molecular data. Although the majority of cultivars have cpDNA from H. albomarginata, the leaf morphology was more similar to that of H. sieboldiana. Using cpDNA results alone, it remains unclear why the majority of cultivars display the H. sieboldiana morphology. We could consider three possibilities. First, it is possible that these cultivars were generated through the artificial hybridization between H. sieboldiana and H. albomarginata, although there are few reports confirming that artificial interspecific hybridization has been practiced in Hosta cultivar formation. Second, it is possible that these cultivars hybridized with natural populations during their domestication. Lastly, we could consider the possibility that H. sieboldiana and H. albomarginata hybridize in nature, as a previous study showed (Takahashi, 2002). Hosta cultivars derived from hybrid origins might exhibit inconsistencies between morphological and molecular data, which were shown in this study.
In this study, because we investigated only cpDNA variations, our results reflected the genetic types of the maternal lineages. It is difficult to elucidate the exact origins of the cultivars. Some studies analyzing both cpDNA and nuclear DNA in the cultivars precisely elucidated their origins (Li et al., 2013; Renner et al., 2007; Roullier et al., 2013). Therefore, nuclear DNA variations should be investigated to determine the exact origins of the cultivated Hosta species in a future study. Such data may clarify the domestication process of the Hosta cultivars in Japan in more detail and explain the inconsistencies between morphological and cpDNA data. Our ongoing study using nuclear DNA data will shed light on these two points.
We would like to thank to Dr. Yonekura for his constructive comments.
